This study investigated a video-based self-evaluation package focused on Head Start teachers' use of praise. Fifty-one Head Start teachers were randomly assigned to one of three groups: immediate self-evaluation, delayed self-evaluation, or a control. The dependent measures were frequency of general and specific praise statements. The independent variables were a treatment package that consisted of observation skills training and self-evaluation (i.e., self-monitoring and goal setting), and the timing of self-evaluation. Three 5- to 10-minute instructional segments, on 3 consecutive days, were taped for each participant. Instructional Segments 1 and 2 were viewed by the teachers in the two experimental groups. The timing of the video feedback and self-evaluation was manipulated. Trained data collectors recorded the frequency of behaviors during the instructional segments. The frequency of praise for the control group remained low throughout the study. Teachers significantly increased the frequency of their general praise statements, after training and after self-evaluation. The frequency of specific praise was significantly higher only after completion of the self-evaluation package (training and self-evaluation). Specific praise may require more skill by the teacher to identify and use it. No effect of feedback timing was observed in the two experimental groups. (Contains 2 tables and 2 figures.)