A collaborative study was carried out which compared the official chemical method, 43.B14-43.B24, the official rat bioassay, 43.165, and the high pressure liquid chromatographic method for vitamin D3 resin, vitamin D3 resin in oil, and dry concentrate. A total of 340 samples were distributed to 17 collaborators for analysis. Five laboratories performed both the chemical and HPLC methods on 5 sets of blind duplicates. A 2-way analysis of variance comparing both methods for each sample showed a significant (P less than 0.01) difference between methods only for Sample 5. When the 2 methods were compared over all the samples, no significant (P less than 0.05) difference was found. Except for Sample 5, there were no differences in the repeatability of the methods. Per cent recoveries on Sample 3, which contained exactly 0.200 X 10(6) IU/g, showed 98.2% for the chemical method and 100.6% for the HPLC method for the 5 laboratories that performed both methods. The assay results of the HPLC and chemical methods are in good agreement with those found by the biological assay on Samples 1-4, but not for Sample 5. Evidence indicates that Sample 5 degraded partially to isotachysterol, and while the HPLC method yielded a reasonable value on this material, the chemical method erroneously showed full potency. An amendment is included for the collaboratively studied HPLC method which detects and eliminates 5,6-trans-vitamin D3, a possible interferant.