How often should you press the RCD test button?

More often than most sites do. Six-monthly is a common baseline for fixed RCDs in a dry, benign environment, quarterly is better, and portable RCDs on construction sites are commonly tested daily or before each use. It takes seconds and requires no tools.

How often do RCDs need a proper trip-time test?

It depends on environment and criticality rather than a single national rule. Annually alongside other compliance testing suits most commercial buildings; harsher environments and safety-critical circuits justify more often. What matters is that the interval is set deliberately and written down.

Is there a legal interval for RCD testing in New Zealand?

The regulations set performance requirements — what an RCD must achieve — rather than a universal testing frequency for in-service devices. The interval is yours to set and justify, in the same way testing intervals are under AS/NZS 3760.

Who can do RCD trip-time testing?

Someone competent with a calibrated instrument who can interpret the results. Work on the installation itself, and any repair or replacement the testing leads to, is electrical work requiring a registered electrical worker.

How often should RCDs be tested in New Zealand?

Q: How often should RCDs be tested in New Zealand?

There are two tests on two intervals. Push-button testing is frequent and done by users — commonly daily for portable RCDs on construction sites, and six-monthly or better for fixed RCDs in benign environments. Measured trip-time testing is periodic, needs an instrument, and is set by environment and criticality.

Two tests, two intervals

The reason this question produces contradictory answers online is that people are answering about two different tests without saying which.

Push-button testing is frequent, free, done by whoever is on site, and takes seconds. Its value is entirely in how often it happens.

Trip-time testing is periodic, requires a calibrated instrument and a competent person, and produces numbers. Its value is in the measurement and the trend.

Neither substitutes for the other, and the difference is explained properly in push-button vs trip-time RCD testing.

Push-button intervals

The honest answer here is "more often than you are doing it", because on most sites the number is zero.

Working intervals:

Situation	Push-button interval
Portable RCDs on construction and demolition sites	Daily, or before each use
Portable RCDs generally	Before each use
Fixed RCDs — outdoor, wet, wash-down or industrial areas	Monthly
Fixed RCDs — workshops, kitchens, plant rooms	Quarterly
Fixed RCDs — dry offices and benign environments	Six-monthly as a baseline, quarterly is better
Any RCD after a significant event	Immediately — flooding, a lightning strike, a major fault, or any electrical work on the board

Three practical points that determine whether this actually happens.

Somebody has to own it. A named person, in a calendar, with a backup for leave. "The maintenance team" does not press buttons.

Warn people what goes off. The single biggest reason push-button testing stops is that somebody tripped the circuit feeding a server, a chiller or a process line during working hours. Map what each RCD feeds before you start, and pick your moment.

Tell them what a failure looks like. An RCD that does not trip on the button is a fault to be reported and acted on, not a device to press harder or write off as sticky. It also should not be reset and forgotten.

If an RCD does not trip when the test button is pressed, treat the circuits it protects as unprotected until a registered electrical worker has looked at it. This is not a "book it in next month" fault.

Trip-time testing intervals

No New Zealand regulation sets a universal interval for in-service RCD trip-time testing. The regulations set what an RCD must achieve — see when are RCDs required — and leave the how to you, in exactly the way the Health and Safety at Work Act leaves the method of managing any risk to the PCBU.

So you set the interval and you justify it. Reasonable starting points:

Environment	Trip-time testing
Construction and demolition sites	Aligned with the three-monthly equipment cycle
Hire fleets and equipment leaving your control	Before hire, and periodically in the fleet cycle
Wet areas, wash-down, commercial kitchens, laundries	Six-monthly
Workshops, factories, yards, external supply	Six to twelve months
Offices, retail, schools — general circuits	Annually, alongside other compliance testing
Safety-critical or high-consequence circuits	Annually as a floor, more often by assessment
Any RCD after flooding, a lightning strike or board work	Immediately

Annual testing alongside the appliance round suits most commercial buildings, and there is a practical reason beyond cost: doing them together means one register, one report and one conversation, rather than three suppliers each holding a fragment of the picture.

How to set your own interval defensibly

The same logic that applies to appliance testing applies here, and the reasoning is what makes it defensible rather than arbitrary.

List the devices and what they protect. Most boards have no schedule. This step alone frequently uncovers circuits nobody knew were unprotected.
Group by environment and consequence. Not by board, and not by building. A single board can feed a dry office and an external wash bay.
Set a base interval per group from the table above.
Shorten for the things that argue for it — outdoor exposure, moisture, vibration, equipment used by people you do not supervise, circuits where a failure has high consequence.
Write down why. One short paragraph per group. This is the risk assessment, and it converts a number into a position you can defend.
Review annually, or whenever the loads change. Adding EV chargers, solar or drives changes the picture — see RCD types and the 2026 changes.

What to record

Per device, per test:

Device identifier and the circuits it protects
Type (AC, A, F, B) and rated residual current
Measured trip times in milliseconds at each test current used
The limit measured against, and pass or fail derived from the numbers
Date, tester, and the instrument used with its calibration status
Push-button test log — dates and results, kept separately and clearly labelled as button tests
The previous readings, so the trend is visible

That last line is the one that turns a compliance record into a maintenance tool. An RCD drifting upward year on year is telling you it is going to fail. That is only visible if somebody kept the numbers.

Ask any provider for the actual milliseconds. The instrument recorded them. A report that says "pass" without numbers is a receipt for a visit, not evidence of a condition.

Where sites get this wrong

Four patterns, in order of how often we see them.

Nobody presses the buttons at all. The most common finding by a distance. The devices have been in the board for years, untouched, and the first time anyone tests one is when we do.

Push-button results are recorded as though they were tests. A compliance file full of "RCD tested — OK" entries containing no measurements. It looks like a record. It demonstrates nothing.

Only the board is tested, and portable RCDs are missed entirely. The portable devices go outside, get dropped, get rained on and are used precisely where the risk is highest. They are also the ones nobody owns. See portable RCDs and PSOAs.

Testing stops after nuisance tripping. An RCD trips repeatedly, someone gets tired of it, and the underlying fault is never diagnosed — sometimes the device is simply bypassed. That is the worst outcome available and it always starts with a legitimate frustration. The diagnosis is usually straightforward: why an RCD keeps tripping.