The authors examined peer coaching in the context of the Mentored Implementation Program developed in the Appalachian Mathematics and Science Partnership. The experimental design contained 6 teachers receiving peer coaching with their 202 students and 5 teachers in the control group with their 105 students. Teachers considered peer coaching a positive experience, identifying scheduling and distance as roadblocks. Centering around (a) organization of learning, (b) management of classroom, and (c) mathematical content and pedagogy, collaborative interactions during the post-classroom-observation conferences were brief and showed (a) a lack of analysis, (b) a positive tone and support, (c) a proportional pattern of talk, and (d) a lack of depth in discussion. Peer coaching was not associated with any improvement in mathematics achievement of students. (Contains 3 tables and 4 notes.)