This 4 x 4 Latin square experiment evaluated the effect of replacement of wheat bran by graded levels of sugar cane molasses on feed intake, rumen fermentation and nitrogen utilization. Four ruminally fistulated crossbred (Bos taurus x Bos indicus) steers were given a basal diet of grass hay ad libitum and 4.0 kg dry matter of wheat bran day-1 (Diet MO). In the other three diets wheat bran was replaced with 1.0 (Diet ML), 2.0 (Diet MM), or 3.0 kg dry matter of molasses (Diet MH). Urea supplementation was used to make the three diets (ML, MM and MH) isonitrogenous with the basal diet (MO). The voluntary dry matter intake of hay (mean 4.01 kg, standard error of the means 0.128) was not affected significantly (P > 0.05) by the treatments. The apparent digestibility of organic matter was similar between diets. Molasses did not depress significantly (P > 0.05) the average value of rumen pH. Increasing the level of molasses increased mainly the molar proportion of propionate at the expense of acetate (P < 0.05). Apparent digestibility of nitrogen was similar between diets (P > 0.05). Daily excretion of allantoin in the urine was not affected (P > 0.05) by molasses. There was a quadratic (P < 0.05) effect caused by molasses so that with Diets ML and MM the daily excretion of uric acid was higher compared with the two other diets. Total microbial purine derivatives (allantoin plus uric acid) and microbial nitrogen were not significantly (P > 0.05) different between diets.