This meta-analysis demonstrates that the video feedback method has a statistically significant effect on the interaction skills of professionals in a range of contact professions. The aggregate effect, calculated on the basis of 217 experimental comparisons from 33 experimental studies involving a total of 1,058 people, was 0.40 standard deviation (SE = 0.07). The effects of training were greater for programs working with a standard observation form of target skills that were central to the program. Results were more positive for outcome measures that measured positive skills rather than negative ones. In addition, molar outcome measures, which were obtained by means of an assessment scale, showed larger effects than micromeasures, which were scored using event sampling. Finally, recommendations are made for video feedback design and for future research. (Contains 4 tables.)