Holstein steers were used in two 5 × 5 Latin square experiments to evaluate the sparing of methionine by alternative sources of methyl groups (betaine or choline). Steers were housed in metabolism crates and limit fed a diet high in rumen degradable protein. To increase energy supply, volatile fatty acids were infused into the rumens, and glucose was infused into the abomasum. An amino acid mixture, limiting in methionine, was infused abomasally to ensure that non-sulfur amino acids did not limit protein synthesis. Treatments for Exp. 1 were abomasal infusion of 1) water (control), 2) 2 g/day additional L-methionine, 3) 1.7 g/day Lcysteine, 4) 1.6 g/day betaine, and 5) 1.7 g/day L-cysteine + 1.6 g/day betaine. Treatments for Exp. 2 were abomasal infusion of 1) water (control), 2) 2 g/day additional L-methionine, 3) 8 g/day betaine, 4) 16 g/day betaine, and 5) 8 g/day choline. In both experiments, nitrogen retention increased (P<.05) in response to methionine, demonstrating a deficiency of sulfur amino acids. Responses to cysteine, betaine and choline were small. The low response to cysteine indicates that either the response to methionine is not due to transsulfuration to cysteine, or that cysteine supply does not alter the flux of methionine through transsulfuration. The small responses to betaine and choline suggest that they do not substitute for methionine. Thus, under our experimental conditions, responses to methionine likely were due to a correction of a deficiency of methionine per se rather than of methyl group donors.