In daily life, independent and safe locomotion requires the ability to continuously adjust the locomotor pattern in response to environmental demands, like obstacles in the travel path. Adjustments of leg trajectories in response to an obstacle or target displacement can be as fast as 120 ms in young subjects (Weerdesteyn et al. 2004; Reynolds & Day, 2005). In obstacle avoidance under time pressure, elderly have been reported to be less successful (Chen et al. 1994). At the neuromuscular control level, however, it is not clear whether this is due to delayed onset latencies, reduced amplitudes, or both. In the present study, obstacle avoidance under time pressure was studied in 10 young (24.4 ± 2.8 years; mean ± SD) and 9 older adults (70.8 ± 4.9 years). The participants walked on a treadmill at a speed of 3 km/h. An obstacle was dropped 30 times in front of the left foot in late stance, early swing, and mid-swing (low, medium and high time-pressure, respectively). Muscle activity in response to the obstacle was measured by surface electromyography (EMG) from the left biceps femoris (BF), rectus femoris (RF), tibialis anterior (TA), and medial head of gastrocnemius (GM). Initial response latencies were determined, as well as response magnitudes over the first 50 ms of the response (normalized with respect to muscle activity in the corresponding phase of the step cycle during unperturbed walking). In both young and older adults, a large (14.1±10.3 (mean ± SD) times control activity) initial response was consistently observed in BF. There was a significant main effect of both time-pressure and age on BF onset latency (ANOVA, p=0.011 and p=0.007, respectively), but no significant age x time-pressure interaction (p=0.57). BF onset latencies were on average 105±9 ms in the young and 119±10 ms (means ± SD) in the older adults. Onset latencies decreased with increasing time-pressure. Response magnitudes of the 4 muscles over the first 50 ms were larger in young than in older adults (increase of 78% in BF, 152% in RF, 65% in TA, and 25% in GM; MANOVA, p=0.04). The results of the present study indicate that in obstacle avoidance under time pressure, both increased central processing time, as indicated by delayed onset latencies, and reduced muscle activation rates are likely to contribute to the reported higher failure rates in the elderly (Chen et al. 1994). Similar results have been obtained in stumbling experiments (Schillings et al. 2005). These findings could help to explain the large numbers of obstacle-related falls in the elderly.