The Global Monitoring Plan (GMP) according to article 16 of the Stockholm Convention on Persistent Organic Pollutants (POPs) requires that POP laboratories must be capable – at any time – to analyse samples for POPs within a variation of ±25%. Based on this target error of 25%, a statistical model using z-scores was applied to assess the performance of analytical laboratories for POPs and a number of matrices. Since the second round of these ‘Bi-ennial Global Interlaboratory Assessment on Persistent Organic Pollutants (POPs)’, carried out in 2012/2013, perfluoroalkyl substances (PFASs) have been included into the proficiency tests. The third round was carried out in 2016/2017. The test materials included test solutions of PFASs analytical standards, the abiotic matrices sediment, air (extract) and water and the biotic matrices fish, human milk and human plasma. The number of laboratories submitting results for PFASs remained quite stable (IL2 = 27 laboratories; IL3 = 29), but there was broader geographic distribution observed in IL3: in addition to the laboratories from Asia and the Western Europe/other groups, two laboratories from Africa participated, two from Central-Eastern Europe and one from the Latin American/Caribbean region. Considering that PFASs were introduced for the first time in round 2, the results were good to reasonable compared to those of a number of other POPs included in the same study. However, it shall also be mentioned that for some matrices and PFASs, the number of laboratories submitting results was too small and the results too scattered to derive a consensus value. This was especially true for the PFOS precursor compounds and the air matrix. Also, laboratories struggle with the analysis of the branched PFOS isomers. These interlaboratory assessments on PFASs gave promising results and demonstrated the importance of proficiency tests in an international environment to generate trust in laboratory results. The need to participate regularly in such intercomparison assessments is highlighted. The results show the current level of PFAS analysis, which varies by laboratory and by matrix rather than per geographic region. •Two rounds of interlaboratory assessments for PFASs gave excellent to reasonable performance.•Worldwide, almost 30 PFASs laboratories regularly participate in proficiency tests.•Excellent results for L-PFOS in test solutions, sediment and human milk.•Not much experience recognized on determination of branched PFOS.•Air samples still a problem for PFOS precursors.