Indiana University-Purdue University Indianapolis (IUPUI) / Stroke is a sudden and devastating medical condition. People who experience a stroke tend to have long-term physical limitations including impaired walking as part of the ongoing consequences of stroke. While a variety of rehabilitation interventions have demonstrated efficacy for improving walking after stroke, none of the interventions have emerged as superior, and prior to this study, therapeutic-yoga had not been tested as an intervention to improve walking recovery after stroke. METHODS: This study was a secondary data analysis of group therapeutic-yoga on walking recovery measures including walking speed, walking distance, and spatiotemporal step parameter symmetry. The walking recovery measures were collected as secondary outcomes in a sub-sample (n=12) in a pilot randomized controlled study (n=47) designed to test the efficacy of 8-weeks of group therapeutic-yoga on balance and fear of falling. Participants in the current study completed 12-weeks of group therapeutic yoga with outcome assessments at baseline, 8-weeks, and 12-weeks. The main analysis was repeated measures ANOVA to assess the main effect of time with additional analyses including effect sizes, percent of participants achieving change greater than or equal to minimal detectable change (MDC), and mean change score comparisons between baseline and 8-weeks, 8-weeks and 12-weeks, and baseline and 12-weeks. RESULTS: Twelve people with chronic stroke enrolled in the study with 9 completing the intervention and all 3 assessments. No significant main effect of time was found on any of the variables of interest. Walking distance demonstrated a trend toward significant change (p=0.064) and step length symmetry demonstrated significant change (p=0.05) between baseline and 12-weeks. Several spatiotemporal step parameter symmetry ratios demonstrated small to medium effect sizes with the majority (91%) being a negative effect. CONCLUSION: Twelve weeks of group therapeutic-yoga appears to be feasible in a population of people with chronic stroke. Walking distance and step parameter symmetry should be tested in a larger sample. An improved understanding of the impact, progression, and remediation of walking asymmetry is needed.