The procedures and errors in the formalin-azo method of determining vitamin B1 used for the testing of B1 crystals or injections were statistically examined. It was found that this method gave error of only within 1.7% and was advantageous in that it needed no strict control of the reagents used.
