A growing body of research documents the positive impacts of teacher coaching, but research contrasting the effectiveness of different coaching approaches is limited. This study contrasted paired coaching -- delivered to two teachers simultaneously -- with traditional coaching for individual in-service teachers. We examined the effects of these two approaches on observations of teachers' classroom management practices and student behavioral outcomes, relative to non-coached conditions, over the course of a single school year and at a 1-year follow-up. We also explored the relative time and cost efficiency of the two approaches. Two hundred fifty-two teachers from 18 elementary and middle schools participated in the randomized controlled trial. Hierarchical linear modeling was used to account for repeated measures nested within teachers and teachers within schools. Combined effects indicated improved global ratings by observers of teacher behavior management ([delta] = 0.29) after a multiple comparison adjustment. Paired coaching was less effective than individual coaching at improving some observed student behaviors, although these did not remain significant after multiple comparison adjustments. Neither model demonstrated sustained effects after one year. Although the paired coaching was significantly more efficient for coach time, it represented just a modest overall cost savings per school. Results indicate a need for more research to identify feasible coaching approaches yielding sustainable effects. [This paper was published in "Journal of School Psychology" v92 p346-359 2022.]