Research on teaching effectiveness concerning how teaching behaviors were related to student learning in teaching physical education (PE) has been extensive; however, most of the studies were conducted at K-12 levels. The features and interrelationships between teaching behaviors and learning activities in collegiate PE classes remained unanswered. The purpose of this study was to examine those features and interrelationships at collegiate PE class settings using direct instruction behavior analysis (DIBA) system. Participants included 15 instructors and their students in an urban university located in the East Coast of the U.S.A. Thirty PE lessons (two lessons per instructor) were videotaped and analyzed using DIBA. Interval recording (percentage data) and event recording [rate per minute (RPM) data] techniques were employed for data collection. Multiple regressions were utilized for RPM data analysis. The percentage data showed that the instructors spent 36.0% of the time on Structuring, 25.5% of time on Informing, 12.9% of time on providing Feedback, 11.4% of time on Observing, 6.3% of time on Questioning, 5.5% of time on Praise/Encouragement, .1% of time on Controlling, and 3.3% of time on None-Above; whereas the students spent 46.9% of time on Motor-Engaged, 31.3% of time on Cognitive-Engaged, 13.5% of time on Get-Equipment/ Relocation, 5.6% of time on Response-Preparing, 2.9% of time on Wait for a Turn (WT), and .2% of time on Off-Task. The above features were confirmed by the RPM data as follows: the Structuring was 2.62, the Informing was 2.34, the Observing was 1.10, the Feedback was .95, the Motor-Engaged was 3.60, the Cognitive-Engaged was 2.75, the Get-Equipment/Relocation was .79, and the Controlling, Wait for a Turn, and Off-Task three categories showed very low frequencies. The results of the Multiple regression analysis revealed that the following six correlations were significant (p < .05): Informing vs. Cognitive-Engaged (r = .881); Structuring vs. Motor-Engaged (r = .658); Questioning vs. Cognitive-Engaged (r = .506); Observing vs. Motor-Engaged (r = .463); Informing vs. Response-Preparing (r = .375); and Observing vs. Get-Equipment/Relocation (r = .363). In conclusion, it is suggested that collegiate PE instructors should reduce Get-Equipment/Relocation, Response-Preparing, and Wait for a Turn on students by using less Structuring and Informing teaching behaviors. The findings of this study may serve as a guide for collegiate PE instructors to modify and improve their instructions.