One hundred sixty crossbred steers (538 ± 36 kg BW) were used in an 84-d experiment with a randomized block design to study the effects of wheat or barley grain processed to 2 different indices on growth performance, feeding behavior, carcass characteristics, stress, and temperament of finishing beef cattle. Treatments were a wheat-based diet (88.4% of diet DM; WH) and a barley-based diet (89% of diet DM; BA), processed to an index of either 75% (HI) or 85% (LO) of their original volume weight. Cattle were allocated to 16 feedlot pens (10 animals per pen, 4 pens per treatment), 8 of which were equipped with the GrowSafe system for monitoring feeding behavior. Flight speed, hair, and saliva samples were collected on d 1, 28, 56, and 84 to determine temperament, acute, and chronic stress. All steers were slaughtered at the end of the experiment, and carcass quality was evaluated. Cattle fed WH had a lower (P < 0.05) meal length and frequency of visits per meal and tended (P = 0.10) to have a lower DMI, meal size, and feeding time than those fed BA. The LO processing index increased (P = 0.05) DMI and reduced (P < 0.05) the G:F and the percentage of saleable meat of the carcass compared to HI. There was a trend (P = 0.09) for a grain × processing index interaction, where cattle fed BA-LO had a lower incidence of severe liver abscesses compared with cattle fed other treatments. Cattle fed WH had greater hair cortisol concentrations (P = 0.01) and flight speed (P < 0.01) than those fed BA. There was a trend (P = 0.07) for a grain × processing index interaction, where heifers fed WH-LO had a lower salivary cortisol than those fed other treatments. Results suggest that a LO processing index had a negative effect on feed efficiency and carcass performance and that the WH diet caused a range of effects on feed intake and behavior indicative of steers with greater excitability and chronic stress.