Effects of live yeast (Saccharomyces cerevisiae) in steam-flaked corn-based diets fed to natural-program beef cattle on growth performance, total tract apparent digestibility, carcass characteristics, and feeding behavior were evaluated in a randomized block design experiment. Steers (n = 144; 341 ± 7.03 kg) were blocked by initial BW and assigned randomly to 1 of the 3 treatments (n = 12 pens per treatment with 4 steers per pen). Treatments included the following: 1) control (CTL; no yeast); 2) low yeast (LY; 1.5 g/animal daily [3 × 1010 CFU]); and 3) high yeast (HY; 3.0 g/animal daily [6 × 1010 CFU]). Technologies such as implants, ionophores, and antibiotics were not used, and the steam-flaked corn-based finishing diets were fed to provide ad libitum access to feed. Yeast was included in a cottonseed meal-based premix as 1% of the dietary DM. Spot fecal samples (twice daily for 5 consecutive days) and diets were composited by pen and analyzed for acid insoluble ash to estimate apparent total tract digestibility of nutrients. Cattle were slaughtered on days 183 (4 blocks) and 204 (8 blocks). Dry matter intake (P ≥ 0.29), ADG (P ≥ 0.17), and G:F (P ≥ 0.33) did not differ among treatments. The percentage of Premium Choice (P < 0.01) carcasses increased linearly with increasing yeast inclusion in the diet. A quadratic response was observed for total tract apparent digestibility, in which steers fed LY had greater digestibility (P < 0.01) of DM by 5.4%, OM by 4.8%, NDF by 15.2%, ADF by 20.2%, CP by 6.2%, and ether extract (EE) by 2.5% compared with steers fed CTL. Feeding behavior was not affected (P = 0.28) by treatments. Live yeast improved digestibility of DM, OM, CP, EE, and fiber, without changing feeding behavior and growth performance of natural-program steers fed steam-flaked corn-based finishing diets.