The Declaration of Helsinki is a statement of ethical principles for medical research involving human participants, adopted by the World Medical Association (WMA) in June 1964. Its scope includes research using identifiable human biological material or data. It establishes responsibilities concerning scientific validity, independent ethical review, informed consent, the assessment of risks and benefits, the protection of participants, and the public communication of research results. The current version was adopted in Helsinki in October 2024.
The declaration is one of the most influential instruments in research ethics. It is not an international treaty and does not itself create a universally enforceable legal code. Its authority instead derives from professional commitments and its incorporation into institutional policies, research governance, publication requirements, and some legal and regulatory frameworks. Its successive revisions record changes in the organization of medical research, including the growth of multinational trials, research involving stored materials and data, and demands for fairer participation and access to research benefits.
Origins and successive revisions
The declaration emerged from efforts to define the responsibilities of physicians conducting research after World War II. The 1947 Nuremberg Code, formulated in the judgment of the Doctors’ Trial, had established principles including voluntary consent, scientific justification, and avoidance of unnecessary suffering. The WMA, founded in 1947, developed additional professional statements addressing physicians’ obligations. The Declaration of Geneva, adopted in 1948, reformulated the physician’s professional pledge; the Declaration of Helsinki addressed the more specific problem of research involving human beings.
The original declaration was adopted at the WMA’s eighteenth General Assembly in Helsinki in June 1964. It distinguished clinical research combined with professional care from non-therapeutic research conducted primarily for scientific purposes. This reflected the central position of the physician-investigator and the difficulty of reconciling a duty to treat an individual patient with the production of knowledge that might benefit others. Unlike the Nuremberg Code, the declaration addressed circumstances in which a legally authorized representative could provide permission for someone unable to consent personally. Its framework therefore accommodated research involving children and other people lacking decision-making capacity, subject to protections.
The 1975 revision in Tokyo was particularly important because it introduced explicit independent committee review of research protocols. Further amendments followed in Venice in 1983, Hong Kong in 1989, and Somerset West in 1996. The Edinburgh revision of 2000 substantially reorganized the document and intensified debate about placebo controls and obligations after a trial. The WMA adopted clarifications concerning placebo use in 2002 and post-trial access in 2004. Revisions in Seoul in 2008 and Fortaleza in 2013 strengthened provisions on matters including registration, publication, compensation for research-related harm, and the accessibility of research results.
The October 2024 revision returned the declaration to the city of its original adoption, sixty years later. It replaced the English term “subjects” with “participants” and placed stronger emphasis on meaningful engagement, structural inequities, research in public health emergencies, and environmental sustainability. It also developed the treatment of vulnerability and the relationship between medical research and contemporary data practices. These changes did not replace the declaration’s central protections with a new ethical system; they extended those protections to changing research circumstances. References to the declaration therefore require attention to the version involved: a provision quoted from 2000 or 2013 may differ in wording, scope, or emphasis from the current text.
Ethical authority, scope, and professional responsibility
The declaration belongs to professional ethics, but its intended influence extends beyond individual physicians. Earlier versions were directed principally to doctors while encouraging others involved in medical research to adopt its principles. The 2024 text explicitly addresses the responsibilities of individuals, teams, and organizations involved in research. This broader framing reflects the distribution of research decisions among investigators, commercial sponsors, universities, hospitals, funders, ethics committees, and public authorities. A participant’s protection cannot depend solely on the conduct of the clinician who obtains consent when other actors control the study design, data access, or resources.
Its central ethical commitment is that the rights and interests of individual research participants must not be subordinated to the objective of generating knowledge. Medical progress depends on research, including research involving people, but that dependence does not authorize unlimited experimentation. The declaration connects this position with respect for dignity, self-determination, privacy, confidentiality, health, and other human rights. It also assigns responsibility for participant protection to qualified professionals rather than transferring that responsibility to participants merely because they have agreed to take part.
The declaration is neither a substitute for national law nor simply a description of existing legal obligations. It directs researchers to consider relevant ethical, legal, and regulatory requirements, while maintaining that such requirements must not reduce the protections it establishes. This creates an international ethical benchmark against which local arrangements may be assessed. It does not, however, give an ethics committee authority to disregard applicable legislation or provide a uniform legal remedy when a principle is breached. Legal consequences depend on the jurisdiction and on how the declaration has been incorporated into enforceable rules, contracts, or professional obligations.
Other instruments address overlapping questions. The Belmont Report of 1979 organizes research ethics around respect for persons, beneficence, and justice. The international guidelines of the Council for International Organizations of Medical Sciences provide more detailed guidance on applying ethical principles, particularly in varied social and economic settings. Good Clinical Practice standards specify operational requirements for conducting and documenting clinical trials. These instruments are related but not interchangeable. The declaration states broad ethical duties; operational standards translate some duties into procedures; legislation determines legally binding requirements. A study may satisfy a documentation standard yet still raise ethical questions about its purpose, participant selection, or distribution of benefits.
Scientific validity and the assessment of risks and benefits
Scientific validity is an ethical requirement because participation can expose people to burdens without producing useful knowledge. The declaration requires research involving humans to conform to generally accepted scientific principles and to rest on adequate knowledge of the scientific literature, relevant information, and appropriate laboratory or animal investigations. This does not prescribe one study design for every question. It requires a defensible relationship between the research question, the methods used, and the conclusions the study can reasonably support. An inadequately designed study cannot become ethical simply by obtaining consent.
The research protocol must explain and justify the study’s design and ethical considerations. In a randomized controlled trial, for example, allocation procedures, comparison groups, outcome measures, and the proposed number of participants affect both scientific reliability and participant exposure. A sample too small to answer the stated question may waste participants’ contributions, whereas unnecessary recruitment can expose more people than the question requires. Problems in statistical planning, systematic bias, and unreliable measurements are consequently not merely technical defects. They can undermine the justification for imposing research-related risks.
The declaration requires an assessment of predictable risks and burdens in relation to anticipated benefits to participants and to others affected by the condition under study. Risks include physical injury, psychological distress, loss of privacy, and other harms relevant to the particular investigation. Burdens may include repeated visits, invasive sampling, time away from work, travel, or demanding monitoring procedures. Benefits must not be assumed merely because a study offers a new intervention: the intervention’s effectiveness and safety may be precisely what remains unknown. Payment for participation is likewise distinct from a therapeutic benefit produced by the intervention.
Risk assessment continues throughout the research rather than ending at approval. Investigators must monitor and document risks and take action when the balance changes. A study may require modification or termination if risks become excessive, or if sufficiently conclusive results remove the justification for continued exposure. The declaration also requires appropriate qualifications, training, and supervision for those conducting research. Its 2024 provisions recognize that protection extends to broader research conditions: medical research should avoid or minimize environmental harm and strive for environmental sustainability. These requirements do not eliminate scientific uncertainty. They establish that uncertainty must be investigated through a justified, competently conducted process, with participant protection maintained as new information emerges.
Independent ethical review and research governance
Independent review is one of the declaration’s principal institutional safeguards. Before research begins, its protocol must be submitted to a research ethics committee for consideration, comment, guidance, and approval. Depending on the jurisdiction, such a body may be called an institutional review board, research ethics board, or independent ethics committee. Its task is not limited to checking whether a consent form contains specified phrases. It evaluates the study’s scientific and ethical justification, participant selection, risk management, recruitment procedures, consent arrangements, and provisions for privacy and follow-up.
The declaration requires committees to be independent of the researcher, sponsor, and other inappropriate influences. Independence concerns both membership and the conditions under which decisions are made. A committee needs sufficient authority and resources to assess research rather than merely endorse decisions made elsewhere. Relevant expertise is also necessary: a technically complex trial may require specialized scientific assessment, while ethical judgment requires attention to participants’ circumstances and the local setting. The 2024 revision strengthens expectations concerning committee familiarity with local conditions and inclusion of a member of the general public.
Review continues after initial approval. Researchers must supply monitoring information required by the committee, particularly information about serious adverse events, and cannot change an approved protocol without appropriate consideration and approval. At the study’s conclusion, a final report containing a summary of findings and conclusions must be submitted. These provisions make ethical review a continuing governance relationship rather than a single administrative permission. They also distinguish committee oversight from the separate activities of sponsors, trial monitors, regulators, and bodies established to review accumulating safety data.
A conflict of interest may affect any part of this system. A sponsor may benefit commercially from favorable findings; an investigator may have financial interests or career incentives; an institution may depend on research income. The declaration requires relevant funding, affiliations, and potential conflicts to be addressed in research documentation and disclosed as appropriate. Disclosure makes competing interests visible, but does not by itself resolve them. Independent review, management of conflicting roles, and restrictions on particular activities may also be necessary under applicable arrangements. In multinational research, the 2024 declaration requires ethics review in both sponsoring and host countries. This recognizes that scientific resources and financial control may be located far from the people who bear the immediate burdens of participation.
Informed consent and limits to autonomous authorization
Informed consent is a process through which a person capable of deciding receives relevant information, understands it sufficiently, and voluntarily authorizes participation. Under the declaration, prospective participants must receive information about the research’s aims, methods, anticipated benefits, potential risks and burdens, funding, relevant affiliations, and potential conflicts of interest. They must also be informed about protections and arrangements relevant to their participation, including compensation and treatment for research-related harm. Consent is therefore more than a signature documenting acceptance of an investigator’s invitation.
Communication must take account of the individual’s informational needs and ability to understand. Technical accuracy alone is insufficient if an explanation is inaccessible because of language, literacy, unfamiliar terminology, or the circumstances in which it is delivered. The researcher must ensure that the prospective participant has understood the information before seeking consent. The declaration also requires participants to know that they may refuse participation or withdraw consent without reprisal. Refusal or withdrawal must not adversely affect the patient–physician relationship or the provision of appropriate care.
The relationship between research and treatment creates particular difficulties. A patient may believe that an investigational intervention has been selected solely for personal therapeutic benefit, even when allocation is randomized or dictated by a protocol. This misunderstanding is often termed therapeutic misconception. Dependence on a treating physician can also create pressure to agree. The declaration therefore requires special caution where a prospective participant is in a dependent relationship or may consent under duress; consent should then be sought by an appropriately qualified person independent of that relationship. These safeguards address undue influence without treating every clinical relationship as inherently coercive.
Some participants cannot provide valid consent because of their age, impaired decision-making capacity, or temporary clinical condition. Permission must then be obtained from a legally authorized representative, together with additional protections appropriate to the research. Where the individual can express assent, that assent must be sought, and dissent must be respected. Research involving people unable to consent and offering no prospect of individual benefit is subject to restrictive conditions, including minimal risk and burden. Emergency research without prior consent is permitted only in narrowly specified circumstances, with justification in the protocol, prior ethics committee approval, and consent obtained as soon as possible from the participant or an authorized representative. A useful research question alone does not establish an exception to consent.
Vulnerability, fair inclusion, and community engagement
Vulnerability in research concerns an increased possibility of being wronged or harmed, or of being unable to protect one’s interests under particular conditions. It may arise from individual circumstances, such as impaired decision-making capacity, or from social conditions, such as poverty, discrimination, displacement, and restricted access to health care. The declaration does not treat vulnerability as a reason to disregard a person’s agency. It calls for specifically considered support and protections corresponding to the circumstances that create the vulnerability.
The 2024 revision places particular emphasis on the context-dependent character of vulnerability. A person or community may be vulnerable in one research setting but not another, and the relevant circumstances may change. For example, recruitment through a service on which people depend may produce pressures absent from recruitment through an independent public invitation. Language barriers or insecure social status may also affect the practical ability to ask questions, decline participation, or report harm. Ethical assessment therefore requires more than placing prospective participants into fixed demographic categories.
Protection must be balanced against the harms of exclusion. Systematically excluding groups can produce gaps in knowledge about the effects of treatments they will later receive. Children, pregnant people, older adults, and people with multiple health conditions have often raised difficult questions about the relationship between precaution and fair access to research. The declaration requires appropriate access to participation for groups underrepresented in medical research, while retaining safeguards tailored to risk and capacity. Research involving people in situations of vulnerability must respond to their health needs and priorities, and those populations should stand to benefit from the resulting knowledge, practices, or interventions. Inclusion is not justified merely because recruitment is convenient.
Meaningful engagement with potential and enrolled participants and their communities is another emphasis of the 2024 text. Engagement can inform research priorities, study design, recruitment materials, acceptable burdens, and communication of findings. It is not equivalent to individual consent: a community representative cannot ordinarily authorize participation on behalf of competent adults. Nor does consultation guarantee agreement on every issue. Its ethical significance lies in allowing research decisions to be informed by the people affected, rather than treating them solely as sources of data. This approach also addresses distributive justice in international research, where the populations bearing study burdens may otherwise have little influence over the knowledge produced or access to its benefits.
Control groups, placebo use, and standards of care
The declaration’s provisions on control groups address a central conflict between methodological clarity and obligations to participants. A placebo can help distinguish the effects of an intervention from changes associated with expectations, clinical attention, the natural course of illness, and other influences. However, assigning a participant to placebo may also mean withholding an effective treatment. The relevant ethical question is therefore not whether placebo-controlled trials are scientifically useful in general, but whether a particular comparison is justified given available interventions and the consequences of not providing them.
The declaration’s general standard is that the benefits, risks, burdens, and effectiveness of a new intervention should be tested against those of the best proven intervention. It recognizes an exception where no proven intervention exists, in which case placebo or no intervention may be acceptable. It also allows another exception where compelling and scientifically sound methodological reasons make a comparison involving placebo, no intervention, or an intervention less effective than the best proven one necessary to determine efficacy or safety. In that situation, participants must not face additional risks of serious or irreversible harm because they do not receive the best proven intervention.
These conditions are cumulative, not alternative justifications. Scientific convenience, lower cost, or a simpler route to statistical significance does not by itself satisfy them. The seriousness and reversibility of foreseeable harm matter, as do possible ways of limiting exposure. Some studies use an investigational treatment or placebo in addition to established background therapy, rather than replacing that therapy. Others incorporate rescue treatment, close monitoring, or early withdrawal criteria. Such measures affect the ethical assessment but do not automatically make a design acceptable; the complete protocol still requires independent evaluation.
Debates became particularly prominent during the 1990s over trials seeking less costly regimens to reduce mother-to-child transmission of HIV in resource-constrained settings. The controversy concerned whether the comparator should reflect an internationally established effective intervention or the treatment ordinarily accessible in the host setting. It exposed a broader problem: local scarcity can make research feasible while also creating conditions in which participants lack meaningful alternatives. The 2000 revision and the 2002 clarification were important stages in the WMA’s response. Subsequent formulations retained limited exceptions while emphasizing protection against serious or irreversible harm. The issue remains distinct from whether a health system can afford an intervention after the trial; that question concerns access and research justice, not solely the validity of the control group.
Research-related harm and responsibilities after a trial
The declaration treats responsibilities to participants as extending beyond recruitment and data collection. It requires appropriate treatment and compensation for people harmed as a result of research participation. This obligation differs from ordinary payment for participation, reimbursement of travel expenses, or access to an experimental intervention. It concerns injury attributable to research and requires arrangements that can function if harm occurs. The declaration does not itself create a global compensation fund or prescribe a single insurance model; legal rules and institutional mechanisms differ across jurisdictions.
Post-trial provisions address a separate problem. An intervention may prove beneficial during a trial, yet participants may lose access when the study ends. This is especially consequential when a trial supplies treatment unavailable through the ordinary health system. Earlier revisions made post-trial access increasingly explicit, and the 2024 text requires sponsors and researchers to arrange provisions in advance for participants who still need an intervention identified in the trial as beneficial and reasonably safe. Exceptions require approval by a research ethics committee. The planned provisions must be disclosed during the consent process.
Advance planning matters because responsibility can otherwise be shifted between sponsors, investigators, host institutions, and governments after the research has generated its results. Relevant questions include what will be supplied, to whom, for how long, and through which delivery arrangements. Continued supply of an investigational product may also depend on regulatory permissions and adequate safety monitoring. A declaration-level duty establishes the ethical requirement, but practical fulfillment requires arrangements suited to the intervention and setting. Post-trial access is not a promise that every experimental treatment will work, remain available indefinitely, or become part of routine care.
The declaration separately addresses unproven interventions used in clinical practice when established options are inadequate or ineffective and participation in a clinical trial is not possible. Such use requires expert advice, attention to possible risks, burdens, and benefits, and informed consent. Relevant information should be recorded and shared appropriately, and the intervention should subsequently be investigated systematically. This provision is not a general exemption from research safeguards and must not be used to circumvent them. The distinction is between a carefully justified response to an individual clinical situation and a systematic investigation intended to establish safety or effectiveness. Neither the seriousness of a patient’s illness nor professional confidence in an innovation removes the need for ethical scrutiny.
Biological materials, data, and privacy
The declaration applies not only to research interventions performed directly on participants but also to research using identifiable human material or data. Blood samples, tissue, genetic information, medical records, and linked datasets can support important investigations without another clinical encounter. Their use nevertheless raises questions about consent, confidentiality, secondary purposes, governance, and the possibility that individuals or communities may be harmed by disclosure or interpretation. A sample’s physical separation from the person who supplied it does not remove every ethical obligation associated with its use.
The 2024 text addresses the collection, processing, storage, and foreseeable secondary use of biological material and identifiable or re-identifiable data. Consent remains central, and collection and storage for multiple or indefinite uses must meet relevant requirements of the WMA’s Declaration of Taipei, revised in 2016. That declaration specifically concerns ethical considerations regarding health databases and biobanks. The relationship between the two instruments allows Helsinki to state general research principles while referring to a more specialized framework for the governance of retained materials and information.
Secondary research can create situations in which obtaining consent is impossible or impracticable. The Declaration of Helsinki does not treat this difficulty as an automatic authorization. Such research requires consideration and approval by a research ethics committee. The distinction between identifiable, coded, and effectively anonymized information is important, but it is not always technically stable. Genetic information and combinations of demographic or clinical variables may permit re-identification when linked to other sources. A claim that names have been removed is therefore not equivalent to a complete assessment of data privacy.
Contemporary computational methods make these questions relevant to research using artificial intelligence, including systems developed from large clinical datasets. Ethical evaluation may need to consider the origins of training data, permissions for reuse, security, potential re-identification, and the distribution of benefits and harms. The declaration does not provide a technical standard for every algorithm or database architecture. Its contribution is to maintain duties of consent, confidentiality, scientific justification, and independent oversight as research moves beyond face-to-face intervention. Compliance with a data-protection statute and ethical acceptability are related but distinct judgments: a legally permitted use can still require scrutiny of whether it respects the conditions under which material was obtained and the interests of those represented in it.
Registration, publication, and public accountability
The declaration connects obligations to individual participants with obligations to the public record of research. Every research study involving human participants must be registered in a publicly accessible database before recruitment of the first participant. Prospective registration creates a dated account of the study’s existence and intended purposes. In trials, it can help identify discrepancies between planned outcomes and those later emphasized in publications. Registration does not establish that a study is ethical or scientifically sound, but it makes aspects of its conduct visible to people outside the research team.
Researchers, authors, sponsors, editors, and publishers have ethical obligations concerning publication and dissemination. Results must be made publicly available, including negative and inconclusive findings as well as positive findings. Funding sources, institutional affiliations, and conflicts of interest must be declared. These duties respond to publication bias, in which findings that appear favorable or statistically significant are more likely to enter the accessible literature. If unfavorable studies remain unavailable, later investigators and clinicians may overestimate an intervention’s effectiveness or underestimate its risks, even when each published report accurately describes its own data.
Transparency also supports reproducibility and the evaluation of evidence across studies. The declaration is not a comprehensive specification for open data, software release, or statistical reporting, and its publication provisions do not override participant confidentiality. It nevertheless establishes that research has an ethical destination beyond possession of results by a sponsor or investigator. Participants may accept burdens on the understanding that their contribution will generate knowledge. Failure to disseminate results can defeat that purpose and lead to unnecessary duplication of studies involving further participants.
The declaration’s practical influence is visible in ethics application forms, institutional requirements, journal policies, and professional guidance. Such use does not mean that all bodies employ the same version or interpret every provision identically. In the United States, for example, a 2008 Food and Drug Administration rule replaced the declaration as the specified ethical reference for certain foreign clinical studies not conducted under an investigational new drug application, using Good Clinical Practice requirements instead. This illustrates the distinction between ethical influence and regulatory incorporation. Public health emergencies present another test of implementation: the 2024 text recognizes the urgency of producing knowledge during such events while maintaining that the declaration’s ethical principles continue to apply. Accelerated review and rapid recruitment do not, by themselves, remove duties of scientific validity, participant protection, registration, and disclosure of results.