This study examined the effect of co-teaching on pre-service teachers' performance in and attitude to Statistics in Research Methods course in Primary education. The pre-test, post-test control group quasi experimental design using 2×2 factorial matrix was adopted as a research design for the study. The subjects comprised 293 pre-service primary teachers. Two public college of Education in South West, Nigeria were randomly selected and randomly assigned treatment. Two stimuli packages and two instruments were developed and used for the study. The packages were Co-teaching Instructional Package (CIP) and Conventional Method Package (CMP). The response instruments were Research Methods Test (RMT) and Attitude Scale (AS).The study tested six hypotheses at 0.05 significant level. The data collected were analyzed using Analysis of covariance (ANCOVA). The independent variable is instructional method while the dependent variable was performance in research methods. The result of the findings showed that there was a significant effect of treatment (instructional methods) on preservice teachers' performance in research methods course [F (1, 288) =277.935 P< 0.05]. The coteaching instructional method group have the higher mean score (X=24.96) while that of the solo teaching group (control) have the least mean score (X=9.81) in performance in statistics aspect of research methods course. The study showed no significant main effect of gender on pre-service teachers' performance in statistics aspect of research methods course. The findings also implied that the interaction of treatment and gender do not hinder pre-service teachers' performance in statistics aspect of research methods course. The paper suggests that co-teaching method should be adopted for teaching any Mathematics related subject as a means of improving students' performance in the subject.