RAMS · Guide · 7 min read

Why your RAMS came back, and how to stop that happening

Ask the people who review RAMS for a living and the same rejection reasons come up every time. Here they are, from the review desk's side of the table, with the fix for each.

By Connor McKay, founder of Trade PA · 3 November 2026

A rejected RAMS costs you twice: the rewrite, and the start date sliding while it goes round again. The good news is that rejection reasons are boringly consistent, because the reviewer is answering one question: does this document prove these people have thought about this job on this site? Everything below is a way of failing that test.

1. The document is visibly generic

The site address is right and nothing else is. The hazards mention plant that is not on this job, the method statement describes a new build when this is a refurb, and paragraph four still says another firm's name because that is whose template it originally was. Reviewers see dozens of these and reject them on smell alone, fairly, because a copy-paste RAMS is evidence of no planning, which is the exact thing the document exists to prove. Fix: start from a decent skeleton (ours is free) and force the specifics: this access, this ceiling height, these occupants, this weather exposure. Ten true sentences beat ten pages of boilerplate.

2. Controls that do not answer their hazards

"Work at height. Control: operatives will take care and wear appropriate PPE." That sentence gets documents rejected on its own, because it is not a control, it is a hope. The reviewer wants the hierarchy visible: can the height work be avoided, if not what collective protection (tower, podium, edge protection), only then PPE, and what specifically. Same for isolation ("safe isolation to HSG85 with lock-off and proving unit" versus "power will be off"), dust ("on-tool extraction and FFP3" versus "dust will be minimised"), and manual handling. Fix: for every hazard, write what a stranger would physically see you doing about it on site. If it is not observable, it is not a control.

3. A method statement that cannot be walked

Reviewers read the sequence like a film: arrive, sign in, isolate, protect, strip, first fix, and so on. Rejections come when the film breaks: testing appears before isolation, materials arrive at floors with no access plan, two trades occupy the same space with no coordination line, or the whole job is one paragraph called "carry out the works". Fix: number the steps and walk them in your head from the van door to handover. Anywhere you cannot picture the next step, neither can the reviewer.

4. Competence and roles left vague

Who is the supervisor on site? Who holds the tickets for the MEWP, the gas work, the electrical testing? "All operatives are experienced" answers neither question. On CDM jobs the PC is obliged to check capability, so a RAMS with no named competence is asking them to do your half of the paperwork. Fix: name the lead, list qualifications and cards against the tasks that need them, and state the supervision arrangement in one line.

5. Emergency arrangements that assume nothing goes wrong

No first aider named, no nearest A&E, no rescue plan for the work at height that the same document lists as a hazard. Reviewers notice the mismatch instantly. Fix: the emergency section is five lines and the fastest section to complete honestly. Complete it honestly.

6. Stale versions and missing signatures

The scope changed on Tuesday and the RAMS still describes last month's job; or the document is fine and the briefing record is blank, so there is no evidence anyone who will do the work has read it. Fix: version and date the document, and treat the sign-off sheet as part of the job, not decoration. A reviewer who sees revision history and signatures relaxes, because it signals a firm where the RAMS is used rather than filed.

The pattern behind all six

Every rejection above is the same rejection: the document failed to demonstrate thinking. Which suggests the honest workflow: think about the job (you were doing that anyway), then let the document capture the thinking with as little friction as possible. That is the design brief behind Eve's RAMS drafting in Trade PA: she builds the draft from your actual job details, and you do the part that was always yours, checking it against the real site and signing it. The AI takes the typing; the competence and the judgement stay with the person whose name is on the sign-off, which is precisely what the reviewer wants to see.