We compared the effects of 2 observer-training procedures. In vivo training involved practice during actual treatment sessions. Video training involved practice while watching progressively more complex simulations. Fifty-nine undergraduate students entered 1 of the 2 training conditions sequentially according to an ABABAB design. Results showed that the 2 training methods produced almost identical scores on a posttraining observational test; however, the video method required fewer training sessions to complete. (Contains 1 table, 1 figure, and 1 footnote.)