A series of experiments with rainbow trout (Oncorhynchus mykiss) were conducted to quantify the dietary requirement for choline as well as to determine whether betaine or methionine could spare the trout's choline requirement. Fish weighing 1.4 and 3.2 g, which were acclimated to a choline-deficient semipurified diet for 14 days before the start of the experiments, were shown to have a choline requirement of 813 and 714 ppm, respectively. When an excess of betaine was added to the choline-deficient basal diet, the choline requirement was reduced to about 400 ppm. These results suggest that about half of the trout's choline requirement must be met by choline itself, whereas the other half can be supplied by betaine. However, the requirement for choline itself must first be met before a response to betaine can be obtained. The addition of betaine to a choline-deficient diet resulted in no improvement in weight gain. Likewise, the addition of L-methionine, isomethyl to 1000 ppm choline, was without effect on growth in a basal diet marginally adequate in sulfur amino acids — suggesting that the trout's requirement for choline is not replaceable by methionine. Because the effects of dietary choline on liver lipid were variable and difficult to interpret, it was demonstrated that growth was more sensitive than liver lipid as an indicator of choline status.