Nine hypotheses concerning the immediate and long-term effects of glare were tested in a fixed-base truck simulator. Eight subjects completed two 4-hr sessions. During one session, simulated headlight glare from following vehicles was presented intermittently in the cab's side mirrors. Designated targets, including pedestrians and vehicles in the mirrors, were presented at the beginning, middle, end, or 5 sec following glare episodes. Response times and accuracy for detecting targets were recorded. Target detection in the presence of glare was slower than in the absence of glare. Pedestrians appearing in the middle of a glare episode were detected faster than those appearing at glare onset, indicating adaptation to glare. Target detection following a 5-sec recovery period was faster than for targets appearing immediately following the glare episode. A longer recovery period did not provide an additional benefit. Glare episode duration did not affect target-detection speed following the glare. Performance did not deteriorate over the course of the glare session, indicating no cumulative effect of glare on target detection. Overall, the impairment effects associated with glare were not as strong and consistent as in previous laboratory studies. Differences in attentional demands between laboratory studies and driving were used to explain these differences.