A collaborative trial comparing New Zealand (NZ) routine methods for the determination of moisture, solids-not-fat (SNF) and fat (by difference) contents in butter with IDF Standard 80 was conducted. A comparison of IDF Standard 10 for the determination of moisture was also included. IDF Standard 10, on average, yielded lower moisture results (0.05%) compared with IDF Standard 80. No further bias was detected when the routine and reference methods were compared for any of the components moisture, SNF and fat (by difference). Repeatability values for moisture found in this trial were in close agreement with those quoted by all the respective methods. The SNF repeatability found for the NZ routine method compared well with that quoted by the method, whereas that found for IDF Standard 80 was significantly larger than that quoted by that method. Accordingly, the repeatability for fat (by difference) was larger than that calculated from the quoted moisture and SNF repeatabilities. Reproducibility values for all methods are reported. The results found in this study must be viewed as preliminary. The presence, or lack, of significant bias between the various methods as well as more precise estimates for repeatability and reproducibility should be established by a larger international collaborative trial. (C) 1998 Elsevier Science Ltd. All rights reserved.