This study focused on teacher behaviors in tennis classes to determine which ones seemed most effective. The study was conducted during a sports camp at the university level. Subjects were 8 tennis instructors who each taught 5 tennis lessons of 3 hours to a class of 6 university students with no experience in any kind of tennis as defined for our purpose. The students were pretested on the defined tasks before the instruction and posttested during the last tennis lesson. Pretest and posttest performances and instructional sessions were videotaped. The behavioral data were collected by coding the videotape recordings. Classes were then grouped in three more effective and three less effective ones, based first on students’ measures of technique learning and second on their evaluation of the teacher. Using the Mann-Whitney U test, it became apparent that the two groups differed significantly on useful time, receiving information, and the time devoted to specific feedback. Thus, it might be concluded that teacher effectiveness in producing learning gains depends upon ability to transform class time into useful time and information time. According to the students, the amount of specific feedback is the most important characteristic of good teaching.