This study aimed to determine a convenient method, applicable in a field setting, to evaluate starting technique for competitive swimmers. Five college swimmers volunteered as subjects for this study. This method calculated the swimming velocity at 50 cm intervals from video footage of the starting phase; that is from the start point to 15 m out. Reference points were marked on a course rope at 50 cm intervals. Video recording using a digital video camera (1/30 sec for a frame) was conducted traveling alongside the swimmer from the poolside. After the video recording, the time each reference point was passed and the swimming velocity for each 50 cm interval were calculated from the video footage. An autocorrelation coefficient was used to examine the intra-tester reliability of the passing time for each of the reference points calculated from the video footage, and the reproducibility of the starting performance of each swimmer. Autocorrelation coefficients for intra-tester reliability were high (0.79 to 0.98 with a lag of 0), as were those for reproducibility (0.84 to 0.93 with a lag of 0). The swimming velocity curve obtained using this method was considered to reflect the characteristics of the changes of swimming velocity caused by the swimmer's actions. Considering its convenience in a field setting, this study demonstrated the method's effectiveness.