Twelve ruminally cannulated steers (average initial BW 357 kg) were allotted to four treatments (three steers per treatment) in a replicated 4 x 4 Latin square design with 21-d periods (12 d for adaptation and 9 d for collection) to compare the effects of protein supplements that differed in percentage of CP and feeding level on low-quality forage utilization. Treatments were 1) control (C), ad libitum access to 5.6% CP prairie hay, 2) C +600 g of DM.steer-1.d-1 of a 43% CP supplement based on cottonseed meal (PS), 3) C + 1,200 g of DM.steer-1.d-1 of a 22% CP supplement based on corn grain and cottonseed meal (GS), and 4) C + 600 g of DM.steer-1.d-1 of a 22% CP supplement based on corn grain and cottonseed meal (LS). Ruminal total VFA concentrations were increased 8% (P < .07) by PS vs GS 1 h after supplementation. Among supplemented steers, ruminal acetate (mol/100 mol) was decreased 1.2 mol/100 (P < .03) by GS vs PS and LS; however, supplementation did not affect (P > .10) acetate proportions compared with C. Neither propionate nor butyrate was affected (P > .10) by supplementation, but among supplemented steers, butyrate proportions were 8% greater (P < .03) for GS than for PS and 5% less (P < .10) for LS than for the average of GS and PS. Ruminal pH did not differ (P > 10) among treatments. Ruminal ammonia concentrations were increased 1.4 to 4.8 mg/100 mL (P < .07) by supplementation and typically were less for LS than for PS and GS at most sampling times. Prairie hay DMI (average = 16.3 g/kg BW) was not affected (P > .10) by supplementation. Fluid dilution rate was 8% faster (P < .01) when steers were supplemented than when they were not fed supplement, and fluid dilution rate was increased 4% (P < .04) by GS compared with PS. Particulate digesta passage rate was not affected (P > .10) by treatment, but total tract retention time was decreased (P < .01) 10% by supplementation. Extent and rate of prairie hay NDF digestion in situ were not greatly affected by supplementation, but in situ disappearance of supplement N was 6 to 10 percentage units less (P < .06) for GS than for PS and 2 to 6 percentage units less for LS than for the average of PS and GS supplements. The PS and GS supplements, but not the LS supplement, increased ruminal ammonia, but percentage of CP in supplements and feeding level had minimal effects on forage DMI. At the feeding levels used, both PS and GS should be effective for meeting the protein requirements of cattle consuming low-quality forage.