Existing functional evaluation tools do not accurately reveal the improved function following botulinum toxin A (BTX-A) injection for post-stroke upper limb spasticity. With the aim of developing an alternate method of measuring functional improvement following BTX-A injection, this study tested the feasibility, validity and reliability of video clip analysis performed by the clinicians. Seventy-nine patients administered BTX-A due to post-stroke upper limb spasticity, were retrospectively evaluated using video clip analysis. Pre- and post-injection video clips recorded at 1-month intervals were randomly allocated and sent to three blinded physician evaluators who were asked to choose the one that seemed more improved in terms of hand motion and associated upper limb reaction during gait. The three physicians chose the post-injection video clip as depicting improved hand motion (82.3%, 79.7%, and 72.2%) and associated upper limb reaction during gait (73.4%, 70.9%, and 70.9%). Kappa and intraclass correlation coefficient as a measure of interrater reliability among the three physicians was 0.86 and 0.79 for the hand, and 0.92 and 0.92 for associated upper limb reaction during gait, respectively. The percent overall agreement of the physicians was 78.1% and 71.7% for hand function and associated upper limb reaction, respectively. Retrospective pre- and post-BTX-A injection video clip analyses is a clinically feasible alternative method to evaluate the improvement following BTX-A injection for post-stroke upper limb spasticity, especially in busy clinical practice setting.