PT/ILC - ISO 17025 section 5.9 - Proficiency Test/Inter-Laboratory Comparisons


Trusted Information Resource
This is a question and frustration-venting thread. The question is below.

Measurement Uncertainty (MU) for accredited labs is always a hot (in more ways than one) topic, and in my experience, so is Proficiency Test/Inter-Laboratory Comparisons (PT/ILC) to meet the requirements of ISO/IEC 17025 section 5.9.

Under APLAC (Asia Pacific Laboratory Accreditation Cooperation) rules, PT/ILC is a requirement for accreditation, where it exists. At least one part of the scope must be completed successfully prior to accreditation, and the rest of the scope successfully completed within four years after accreditation.

Now the question. How many labs have had difficulty in obtaining PT/ILC, or have opinions/questions/etc. regarding its value?

Don't hold back. I do ask if you have have difficulty that you mention if you are a test or cal lab, and what the difficulty is.

As an example, until just recently, there was no formal PT for analytical balances, because the instruments cannot be shipped everywhere and expect to get even close results, due to the nature of the instrument.



Am I wrong in thinking the PT/Inter-lab is just one of a number of ways to meet the requirements of 5.9? As long as you have a procedure for monitoring the validity of your tests, you are not required in 17025 to participate in PT. If it is a requirement of APLAC, maybe they have some information supporting the reason why it is required. Does A2LA require it in your area? I can see where it has value to certain labs but it has limited value to others. I guess it is a suggested method in 17025 because it works for some and not for others. If you get anything from APLAC, I'd be interested in seeing it.



Trusted Information Resource
Hi Dave,

Yes, A2LA does in fact require it, as they - like us - are signatories to the APLAC arrangement.

APLAC document PT001 does specify the requirement.

You are correct, PT/ILC is one of the examples, not the requirement as described in the main body of the clause. However, it is the one specified by APLAC. The others can also be included, but if and where PT/ILC is available then it is a requirement, according to APLAC PT001.

Hope this helps.


Ken K

This is the policy we must follow;


Laboratories that wish to become accredited and maintain their accreditation are responsible for participating in a proficiency testing, interlaboratory comparison or a round robin testing program that will meet the requirements of the international accreditation community. At a minimum, the test must meet the requirements of ISO/IEC Guide 43-1.

We maintain a listing of acceptable PT program providers on the website under “Informational Resources”. This listing should be reviewed by the laboratory to determine when changes are made and on a periodic basis. We strongly encourage the use of an “Accredited” PT provider over another, albeit acceptable PT provider simply because Accreditation of the program provides additional confidence to the user of the PT program. It is always permissible however, for any of the PT providers on the list to be used where they have a program that is operated in conformance with Guide 43 for the parameter in question.

Any proficiency test, interlaboratory comparison or round robin (ILC/PT) that does NOT utilize one of these approved PT providers must have the “Proficiency Scheme” approved prior to starting the proficiency test. The scheme should be provided to Operations in the form of a PT/ILC Application Form. The application (Form 28.14) and supporting documents will be judged against the requirements of ISO/IEC Guide 43-1. Any application that does not meet the requirements will not be approved.

Testing Laboratories only: If there are no ILC/PT programs available for Major Field Parameter’s on the laboratory’s accredited scope the laboratory may complete appropriate internal performance evaluations and repeatability studies.

To answer your question Hershal, we are a test lab and have had difficulty finding approved ILC/PT providers for our Major Fields. So now we have to fill out an 11 page application that takes hours to complete, send it in and wait...and wait...and wait for approval/rejection of the scheme.

We have been writing our own ILC/PT schemes per Guide 43-1 using the materials we test. Two other labs in our company participate. We have run six schemes so far and use the Guide 43-1 criteria for calculating Percent Difference and Z-Scores. We do not calculate En scores because we have not been required to list MU (at least not yet :rolleyes:)

Now, my opinion on ILC/PT...ISO 17025 states in 5.9...

The laboratory shall ensure the quality of results by monitoring test and/or calibration results. This monitoring shall be planned and reviewed and may include, but not be limited to, the following;

b) participation in interlaboratory comparison or proficiency testing programmes;

"May" include, not require. So why are we required to do ILC/PT if ISO 17025 does not require us to do so, just suggests we do so? To me any approved ILC/PT by an approved provider is just another added expense on top of our accreditation fees. Those providers do not perform testing on the materials we test every day. So what value do they actually provide?

Gee...another one of those subjects which just make me :mad:

Thanks Hershal... :agree1:

Voodoo Queen

We are a testing laboratory seeking accreditation for our own in-house developed methods and are having trouble finding accredited PT/ILC providers due to the nature/matrices of our products. There is another lab in our organization that we would like to start an ILC program with since they will have access to the same methods as we do, but they are not accredited. I haven't found anything in the standard, nor in the policies of the accrediting body we're working with, that state we must use an accredited provider -- this leads me to believe our plan is okay and justifiable if questioned. Could anyone confirm?


Trusted Information Resource
Before you get too far down the road Voodoo Queen, check with your chosen AB. If you have not selected one then shop around and make this one of the questions.


If you plan just to do your own equipment... do you have to be accredited?


Trusted Information Resource
For only your own equipment, most likely not, but it is a good practice as you then have external validation of the technical proficiency (known in accreditation as competence) and of the soundness of your quality system.

PT, th subject of this particular thread is easily available in pressure, so you would find it is a requirement to achieve accreditation. Not a bad thing though, as it is - simply put - benchmarking of your performance to other similar labs.

Ken K, I agree with you, the Clause verbiage itself (minus the examples) describes what is actually sought. The reason APLAC makes it mandatory - where it exists, as it does not in every field - is because it is benchmarking.

Voodoo Queen, not sure if you have selected an AB yet, but here are questions that can help if you have not:
Will they provide a quote?
Can you discuss the technical parts of your scope with AB Staff, or only with contractors?
Do they return calls and are nice?
Are the price lists on-line and available for everyone to see?

Let us know how it is going.


Hershal, I think MeasurePT now has a PT for analytical balances. Contact Jeff Gust to see.


We are interested in 17025 as a Calibration lab for MTP, parts of the business has already been accredited by A2LA. The issue is that we have been re-calibrating product for years, prior to becoming accredited. We have established labs, where the master units are cross-checked on a weekly basis by the lab. We monitor this with warning values and limits. The master units are calibrated at NPL once/year and yet the accredited lab was also required to carry out proficiency testing (written by the PLC)on a 6 monthly basis, between two intracompany labs. The results do not give us any further information than what we have already, so can we remove this additional testing? With customer products we test before shipping, we could do an extra test once every 3 months in the other lab as confirmation. Would this be acceptable, rather than doing the long winded proficiency test?
Last edited by a moderator:
Top Bottom