Roughage sources were compared in flaked milo-based diets that contained 35% chopped alfalfa hay (AH, control diet) or with cottonseed hulls (CSH) or chopped wheat straw (WS) replacing half the AH. Latin square experiments were used to measure total tract digestion coefficients, particulate passage rates (rare earths), liquid turnover rates (Co-EDTA), and rumination time in six growing steers (Exp. 1) and in situ digestion of DM and NDF, ruminal pH and ruminal DM distribution in three mature, ruminally cannulated steers (Exp. 2). Rates of passage from Exp. 1 and rates and extents of digestion from Exp. 2 were used to calculate apparent extent of ruminal digestion (AED). In Exp. 1, total tract digestibilities of DM and NDF were lower (P less than .05) by 7 and 22%, respectively, when CSH, but not WS, were included in the diet. Digestibility of cell solubles was not different (P greater than .10) among diets. Inclusion of WS increased (P less than .10) rumination time by 36%, and CSH increased intake (P less than .10) by 17% over the control diet. In Exp. 2, there tended to be (P less than .20) increased in situ digestion of milo and AH in the WS diet. Measures of ruminal pH were similar for all diets. The AED for AH and milo DM and NDF, and the proportion of total tract NDF digestion occurring in the rumen (50, 47 and 62% for control, CSH and WS diets, respectively), were highest (P less than .05) for the WS diet. This resulted in similar total tract digestibilities for the WS and AH diets. The two low-digestibility roughages had different effects at this concentrate level; wheat straw enhanced apparent extent of ruminal digestion for NDF of other ingredients in the mixed diets, but cottonseed hulls did not.