Experimental and control groups of 10 hemiparetic stroke patients each underwent a 6 week, twice daily gait training program. The control group participated in a conventional physical therapy gait program. The experimental group trained in the same basic program with the addition of rhythmic auditory stimulation (RAS). Patients entered the study as soon as they could complete 5 strides with hand-held assistance. The training program had to be completed within 3 months of the patients' stroke. In the experimental group RAS was used as a timekeeper to synchronize step patterns and gradually entrain higher stride frequencies. Study groups were equated by gender, lesion site, and age. Motor function was assessed at pretest using Barthel, Fugl-Meyer, and Berg Scales. Walking patterns were assessed during pre- and post-test without RAS present. Pre- vs post-test measures revealed a statistically significant (P<0.05) increase in velocity (164% vs 107%), stride length (88% vs 34%), and reduction in EMG amplitude variability of the gastrocnemius muscle (69% vs 33%) for the RAS-training group compared to the control group. The difference in stride symmetry improvement (32% in the RAS-group vs 16% in the control group) was statistically not significant. The data offer evidence that RAS is an efficient tool to enhance efforts in gait rehabilitation with acute stroke patients.