The Genealogical Proof Standard: Ensuring Accuracy in Your Research

The Genealogical Proof Standard (GPS) is the benchmark framework used by professional genealogists to determine whether a conclusion about family relationships or identity is sufficiently supported by evidence. Developed and maintained by the Board for Certification of Genealogists (BCG), the GPS defines five specific elements that must all be satisfied before a conclusion can be considered "proved." It applies to everything from confirming a birth year to resolving disputed parentage — and understanding its structure explains why genealogical conclusions vary so dramatically in reliability.

• Definition and scope
• Core mechanics or structure
• Causal relationships or drivers
• Classification boundaries
• Tradeoffs and tensions
• Common misconceptions
• Checklist or steps (non-advisory framing)
• Reference table or matrix

Definition and scope

The GPS entered mainstream genealogical practice through the BCG's publication Genealogy Standards, first issued in 2000 and updated in a second edition in 2019. It emerged from the recognition that genealogy, practiced without a rigorous evidentiary framework, produces conclusions ranging from solidly documented to entirely invented — and that the two categories can look identical to the untrained eye.

The standard applies to any genealogical conclusion: a relationship claim, an identity assertion, a date, a place of origin. Its scope is not limited to professional genealogical work. The BCG and the National Genealogical Society (NGS) both encourage amateur researchers to apply it, though the degree of formal documentation varies. The National Genealogical Society has published its own "Standards for Sound Genealogical Research," which align with the GPS's core logic even if they use different language.

The GPS does not rate individuals as good or bad genealogists. It rates conclusions as proved or not proved. That distinction matters: a researcher can be meticulous and still reach an unproved conclusion if the evidence record is genuinely thin.

Core mechanics or structure

The GPS consists of exactly 5 elements, all of which must be satisfied. Satisfying 4 out of 5 does not produce a proved conclusion — it produces a reasonably supported hypothesis, which is a different thing.

1. Reasonably exhaustive search. All sources that a reasonable, competent researcher would know to consult must be searched. "Reasonably exhaustive" is not the same as "every document ever created." It means every source type that is known to exist, survives, and is accessible for the relevant time, place, and population. For a mid-19th century Ohio farmer, that might include US Census records, vital records, land and property records, probate and will records, and church records, among others.

2. Complete and accurate citations. Every source used must be fully cited so that another researcher can locate it independently. The BCG's citation model draws on Elizabeth Shown Mills's Evidence Explained (3rd ed., 2017), which provides citation formats for hundreds of genealogical source types. This is examined in depth at citing genealogical sources.

3. Analysis and correlation of evidence. Evidence from all searched sources must be analyzed individually and then correlated — meaning the researcher identifies agreements, disagreements, and gaps across the full evidence set rather than cherry-picking favorable records.

4. Resolution of conflicting evidence. When sources contradict each other (a birth year verified as 1847 in one record and 1851 in another), the conflict cannot be ignored. The researcher must explain which source is more likely correct, and why, based on source type, informant knowledge, and proximity to the event. The resolving genealogical conflicts framework addresses this directly.

5. A soundly reasoned, written conclusion. The conclusion must be written out, with the reasoning documented. This is not optional. An unwritten conclusion — even one fully supported in the researcher's mind — does not satisfy the GPS because it cannot be independently reviewed or replicated.

Causal relationships or drivers

The GPS exists because genealogical sources are structurally unreliable in ways that aren't always obvious. Death certificates, for example, are created at the moment of death by informants who may have limited knowledge of the deceased's early life. A death certificate provider a birthplace is an original source, but it contains derived information provided by a secondary informant — making it weaker evidence for birthplace than a baptismal record created at the time of birth by someone present. The primary vs. secondary sources framework maps these distinctions.

The GPS is also a response to the specific failure modes of genealogical research: confirmation bias (searching only records that support a preferred lineage), incomplete searching (stopping when one answer is found), and undocumented copying (repeating errors from other researchers without verifying the original source). Without a formal standard, these failure modes propagate silently through family trees, often for generations.

The genealogy industry's move toward online genealogy databases has amplified both the opportunity and the risk. Access to digitized records expanded dramatically after 2000, but so did the volume of unverified content being shared across collaborative platforms. A single error entered into a widely shared tree can fork into thousands of copies within months.

Classification boundaries

The GPS operates within a broader system of source, information, and evidence analysis — often called the "evidence analysis matrix," drawn from Mills's Evidence Explained. Understanding where the GPS sits relative to adjacent concepts prevents category errors.

Source refers to the artifact or document itself (original, derivative, or authored). Information refers to the content within a source (primary or secondary, based on the informant's knowledge). Evidence refers to how that information is used in answering a specific research question (direct, indirect, or negative). The GPS governs how all three categories are assembled into a proved conclusion — it sits above the evidence analysis matrix, not within it.

The GPS also intersects with, but is not identical to, the standards governing proof summaries versus proof arguments. A proof summary is a concise statement appropriate when the evidence is essentially uncontested and the conclusion is straightforward. A proof argument is required when sources conflict or when the evidence is indirect — the reasoning must be laid out fully in narrative form. The BCG's Genealogy Standards (2019) defines both.

Tradeoffs and tensions

The GPS's requirement for a "reasonably exhaustive search" generates the most friction in practice. The standard is intentionally not absolute — requiring every conceivable source would make proof logically impossible, since there is always another archive, another box, another cousin who might have a document. The word "reasonably" does real work here, and reasonable researchers disagree about where the boundary falls.

A second tension lives in the written conclusion requirement. For hobbyists maintaining a family tree rather than writing a formal proof argument, the documentation standard feels burdensome. The BCG acknowledges this implicitly by maintaining separate standards for different research contexts, but the GPS itself does not formally relax the written conclusion requirement.

DNA evidence has introduced a third tension. Genetic genealogy tools can establish biological relationships with high statistical confidence, but that confidence is probabilistic, not certain. A shared autosomal DNA segment of 1,800 centimorgans points almost unambiguously to a first-degree relationship, but the GPS's written conclusion requirement means the researcher must still articulate how that evidence correlates with documentary sources — a task that genetic genealogy tools can support but not replace.

Common misconceptions

"A GPS-compliant conclusion is a guaranteed fact." The GPS produces a "proof" in the legal-argument sense — a conclusion that survives scrutiny given available evidence — not a metaphysical certainty. New evidence can always reopen a proved conclusion.

"Citing a source satisfies the GPS." Citation satisfies element 2, but does nothing for elements 1, 3, 4, or 5. A well-cited conclusion built on a single record fails the GPS on multiple counts.

"The GPS only applies to professional work." The BCG publishes the GPS and tests against it for certification, but the NGS and the genealogy research methods community broadly encourage its application in all research, regardless of professional status.

"Negative evidence doesn't count." Negative evidence — the absence of a name in a record where it would be expected — is a legitimate evidence type within GPS-compliant research. A man absent from every census between 1850 and 1880 in a county where he supposedly lived is a significant finding, not a non-finding.

"Family trees on collaborative platforms constitute evidence." Compiled genealogies — whether in print or digital — are authored works, not primary sources. They carry the errors of their creators. Using them without verification of their underlying sources violates element 1 of the GPS.

Checklist or steps (non-advisory framing)

The following steps reflect the sequential logic of GPS-compliant research as described in the BCG's Genealogy Standards (2019 ed.):

1. Define the research question specifically — name the person, relationship, and event in question before searching begins.
2. Identify all known source types applicable to the time period, geographic location, and population group.
3. Search all identified sources that survive and are accessible, documenting which sources were searched and which were unavailable.
4. Record complete citations for every source at the point of use, following the Evidence Explained citation model.
5. Classify each piece of information as primary or secondary based on informant knowledge.
6. Identify each item of evidence as direct, indirect, or negative relative to the research question.
7. Correlate evidence across all sources, noting agreements and conflicts.
8. Resolve conflicts by evaluating source quality, informant proximity, and internal consistency.
9. Write the conclusion with explicit reasoning connecting the evidence to the answer, in a form reviewable by another researcher.
10. Reassess if new sources become available or new conflicts emerge.

Reference table or matrix

The GPS framework is the backbone of credible genealogical research — and understanding it changes how a researcher reads the genealogyauthority.com home resource. What looks like a settled family history often turns out, under GPS scrutiny, to be a well-organized collection of hypotheses waiting for one more archive box to open.