Investigations concerning the stability of a group of factors contained in cod liver residue and displaying vitamin B12activity have led to the conclusion that under the conditions prevailing during the rendering process, as developed at the Atlantic Fisheries Experimental Station, little or no deactivation occurs. This is in sharp contrast with what occurs under more drastic conditions such as used in the alkali-digestion process.The study of the microbiological method used for the determination of vitamin B12activity has shown that one step of the analytical procedure, namely, that phase during which the extraction of the sample is performed, is responsible for anomalous results. Deactivation occurs during this stage to an extent which varies according to the previous history of the material. The addition of thioglycolic acid or ascorbic acid during this period does not effectively protect the material unless culture medium is added as well. Medium alone is inefficient. Sodium bisulphite on the other hand exerts protection even in the absence of medium. This effect of sodium bisulphite applies not only to cod liver residue, but also to materials of diverse origin, mammalian viscera included. The data obtained show that vitamin B12itself is not involved; presumably vitamin B12aand vitamin B12bare, although there is no direct evidence of their presence in cod liver.As a protection during extraction, 0.08 per cent to 0.1 per cent of sodium bisulphite calculated on the amount of wet material appeared to be optimum.