Hold points vs witness points: what’s the difference?

Published 6 July 2026 · VariationFlow

Hold points and witness points both invite inspection, but only one stops the work. A clear guide for subcontractors: the difference, the S/H/W/R notation, notice periods, what happens if you build past a hold point, and how to manage them without holding up the programme.

Hold points and witness points are two of the most important, and most confused, terms in site quality. Both are moments in an Inspection and Test Plan where someone is meant to check the work. The difference is simple but consequential: a hold point stops the work, and a witness point does not.

The core difference

• A hold point is a mandatory stop. Work must not proceed past it until the nominated party, usually the main contractor or client, has inspected the work to date and signed it off. Build past an unsigned hold point and you are exposed.
• A witness point is an invitation. You notify the other party that an inspection or test is about to happen and give them the chance to attend. If they choose not to, you may proceed and record that they were given notice.

Put plainly: at a hold point the ball is in their court and you wait; at a witness point the ball is in your court and you continue, having offered them the chance to watch.

The S/H/W/R notation

On an ITP you will usually see checkpoints marked with a letter for their type:

• H, Hold point: a stop until signed off.
• W, Witness point: notice given, attendance optional, work may proceed.
• S, Surveillance point: a routine in-process check by your own team.
• R, Review point: a document or record is reviewed rather than a physical inspection.

Notice periods

Both hold and witness points usually carry a required notice period, often 24 or 48 hours, so the inspecting party can arrange to attend. Give too little notice and a genuine attendance can be missed; give none and a witness point is worthless as evidence. The notice, and the fact it was given, is part of the record: who was notified, when, and for what.

What happens if you build past a hold point

This is where hold points earn their name. If you proceed past a hold point without the required sign-off and the work is then covered, a pour goes ahead, a ceiling closes, fire-stopping is boxed in, the inspecting party is within their rights to require you to open it back up to prove it was correct. That is rework, delay and cost, all avoidable. The whole purpose of a hold point is to force the check before the evidence becomes inaccessible.

Managing them without wrecking the programme

The practical tension is real: hold points protect quality but can stall progress while you wait for a signature. Three things keep them from becoming a bottleneck, give notice early and in writing, submit the completed work for sign-off the moment it is ready rather than at the next walk-round, and make it easy for the inspecting party to sign without coming to site.

That last point is where digital quality records change the economics. When a hold point can be submitted from site with its evidence attached, and the main contractor or client can review and sign it from a link on their phone, no site visit, no account, the wait between "work finished" and "signed off" collapses from days to minutes, and the programme keeps moving without anyone building past an open hold point.