Stable 15N was used to evaluate the influence of bacterial contamination on in situ DM and N degradabilities (Dg) of meat and bone meal (MBM), soybean meal (SBM), and wheat bran (WB) in two steers. Bacterial DM and N contamination ranged from 2.4 to 28.6% and 2.1 to 56.8% of residual DM and N, respectively. Effective N degradability increased when bacterial contamination was taken into account (P < .05). The difference was low for MBM (2.4%) and for SBM (3.4%) but high for WB (12.2%). Theoretically, using solid-associated bacteria should give the most accurate correction for bacterial contamination; however, results showed that Dg of N based on liquid-associated bacteria were identical for MBM and SBM (P > .05) and different for WB (P < .05). In a second experiment, five treatments were applied to incubated feeds to remove bacteria fixed to the residues and consequently to determine directly the Dg of DM and N corrected for the bacterial contamination without the need for a marker. These treatments involved chilling for 6 h at 4 degrees C in saline solution alone (T1) or with a commercial detergent (T2), or with sodium dodecyl sulfate (T3) or with methylcellulose (T4), followed by pummeling in a stomacher for 5 min. The last treatment was only machine washing twice (T5). The Dg of DM can be directly determined following the first four treatments, nevertheless their application to MBM and SBM led to higher Dg of N than that corrected for the bacterial contamination determined in the first experiment (P < .05).(ABSTRACT TRUNCATED AT 250 WORDS)