Evidence collection and management: the evidence a programme keeps
Evidence is a claim that can be checked, not a document filed the week before an audit.

Evidence is not a document type. It is a claim that can be checked, and a control assessment produces it by applying assessment methods to assessment objects. The question here is narrower than the quality of one item of evidence and wider than what a single auditor happened to request: how does a programme produce evidence, keep it and reuse it while controls, teams and frameworks change? Its enemy is the dossier assembled the week before an audit — real paperwork, produced for the occasion, with nothing behind it.
What produces evidence
SP 800-53A states the mechanism in one sentence: assessment objectives “are achieved by applying the designated assessment methods to selected assessment objects and compiling or producing the evidence necessary to make the determination associated with each assessment objective” (lines 2160–2162). The methods are three — “examine, interview, and test” (line 1052). The objects fall into four classes — “specifications, mechanisms, activities, and individuals” (line 1039): the document-based artefacts, the hardware and software safeguards, the protection-related actions people perform, and the people themselves. Evidence is produced by pairing a method with an object against a determination the objective already names; it is not scooped up from whatever happens to exist.
the container what fills it
assessment plan (CA-2) controls, procedures, environment, team, roles
methods x objects examine | specifications
interview | mechanisms
test | activities / individuals
depth and coverage how much rigour, how much breadth
ISCM and CA-7 the same pool, produced continuously
Monitor (M-4, M-5) plans, reports and POA&Ms updated, posture reported
Not every method and object, and not the same depth
The catalogue is a menu, and the source says so: “Organizations are not expected to employ all of the assessment methods and assessment objects contained within the assessment procedures” — instead they “have the inherent flexibility to determine the level of effort needed and the assurance required for a particular assessment” (lines 605–609). The record of how much was gathered is a pair of attributes. Depth “addresses the rigor and level of detail in the examination, interview, and testing processes” (lines 1087–1088), with values basic, focused and comprehensive; coverage “addresses the scope or breadth of the examination, interview, and testing” (lines 1090–1091), with the same three values. Two organisations can both say a control was examined. The attributes are what tell a reader whether that sentence means the same thing in both.
The plan is the container
Evidence has a container, and CA-2 builds it. The organisation selects “the appropriate assessor or assessment team for the type of assessment to be conducted” (line 6341) and develops “a control assessment plan that describes the scope of the assessment”, listing the controls and enhancements under assessment, the procedures used to determine effectiveness, and the “assessment environment, assessment team, and assessment roles and responsibilities” (lines 6344–6352). The plan is approved before the assessment runs. SP 800-53A reaches the same place from the assessor’s side: the plan determines which controls are assessed, selects and tailors the procedures and — in the source’s own words — “[assigns] depth and coverage attribute values” (line 1744). The plan is what converts a set of possible evidence into an expected set: the difference between evidence a programme keeps and evidence someone remembers to ask for in time.
Collection is a programme, not an event
An audit is an event; a programme produces evidence continuously. SP 800-137 gives the reason: organisation-wide monitoring “cannot be efficiently achieved through manual processes alone or through automated processes alone” (lines 247–248), and “ISCM is most effective when automated mechanisms are employed where possible for data collection and reporting” (line 345). Where a process stays manual, the requirement is not speed but reproducibility — “Where manual processes are used, the processes are repeatable and verifiable to enable consistent implementation” (lines 248–249). CA-7 makes the same collection ongoing at system level: “Ongoing control assessments in accordance with the continuous monitoring strategy” and “Correlation and analysis of information generated by control assessments and monitoring” (lines 6772, 6777). Assessment and monitoring are two taps into one pool, and the correlation step is what stops them from being two unrelated reports.
Reuse is re-mapping, not re-filing
Evidence produced elsewhere is tempting, and the sources treat it carefully. CA-2’s enhancement for leveraging results from external organisations states that “reusing existing assessment evidence can decrease the time and resources required for assessments” (lines 6497–6498) — but reuse is conditional. SP 800-53A warns that “the content and applicability of existing assessment evidence varies”, so a mixture of methods and objects is usually needed to produce what a determination requires, and organisations must make “appropriate adjustments in assessment procedures to obtain the requisite assessment evidence from external providers” (lines 1836–1842). Before reusing a result, the assessor “determine[s] the credibility of the assessment evidence, the appropriateness of previous analysis, and the applicability of the assessment evidence to current system operating conditions”, and the date and type of the original assessment are documented in the plan and the report (lines 1963–1966). Evidence borrowed for one framework and refiled against another is not reused; it is re-labelled.
Where the evidence goes as monitoring runs
Collection does not end at the report. The Monitor step of the Risk Management Framework gives the pool two duties: M-4 will “Update plans, assessment reports, and plans of action and milestones based on the results of the continuous monitoring process” (line 5321), and M-5 will “Report the security and privacy posture of the system to the authorizing official” on an ongoing basis (line 5384). The evidence pool therefore has a consumer and a refresh cycle, not only a shelf. That is what keeps it current enough to be reused at all.
Level and prerequisites
L4 — programme: this sheet is about the pipeline that turns a determination into evidence, keeps it fed and reuses it across many controls and audits, rather than about one assessment or one artefact. It assumes the L1 vocabulary of control and evidence, the L2 records (control objectives and verification criteria) and the L3 traceability chain from a treated risk to the owner and the evidence. The mapping row from requirement to control to evidence, the quality and freshness of a single item, and the audit engagement that consumes all of it belong to other sheets in this layer; this one owns only what a programme collects, keeps and reuses.
Where to go next
- Knowledge — the Knowledge register this sheet belongs to.
- Cybersecurity Governance — the macro-area.
- Assurance — the node this sheet sits in.
References
- NIST SP 800-53A Rev. 5 — how evidence is produced (assessment objectives, methods, objects), the depth and coverage attributes, the assessment plan, and the reuse of external evidence.
- NIST SP 800-53 Rev. 5 — CA-2 (the assessment plan and its approval) and CA-7 (ongoing assessments, monitoring and their correlation).
- NIST SP 800-137 — the ISCM programme: automation as the aim, repeatable and verifiable manual processes, and the continuous production of evidence.
- NIST SP 800-37 Rev. 2 — the Monitor step: M-4 updating plans, reports and POA&Ms, and M-5 reporting the posture.