Experiments were conducted to test the hypothesis that metabolically labile methyl groups and not methionine per se are limiting the growth of starting broiler chicks fed corn-soy-grease diets formulated to meet NRC (1977) recommendations (except for methionine). The basal diet contained approximately 23% protein, 3200 kcal ME/kg. 32% methionine, .42% cystine, and 1300 ppm choline. In the first two experiments the diets were fed to quadruplicate lots of 10 chicks in battery brooders. Average weight gain (g) and feed per unit gain for supplements of methyl group sources were: 1) methionine: 424, 1.48; 2) choline Cl: 406, 1.49; and 3) betaine HCl: 416, 1.48. These values were not significantly different from one another (P greater than .05) but were significantly different (P less than .05) from the values for the basal (362, 1.57) and from the DL-homocystine supplemented groups (376, 1.56). This indicated that methyl group utilization was not impaired by a lack of the carbon skeleton of methionine. Five 3-week assays with five groups of 10 chicks per treatment (1050 chicks) were conducted to determine the choline requirement using the basal diet without supplemental methionine. Results from the five assays were pooled in computing regressions. Significant regressions (P less than .05) were found for broken-line linear (R2 = .630), quadratic (R2 = .556), and saturation kinetic (R2 = .642) models. The requirement was determined to approximately 2130 ppm by the broken-line linear model. The most economic feeding levels were estimated to be 4100 and 1910 ppm by the quadratic and saturation kinetic models, respectively. The economic implications of the three models are discussed.