College instructors use a variety of approaches to teach students to reason more effectively about issues with a moral dimension and achieve mixed results. This pre-post study of 423 undergraduate students examined the effects of morally explicit and implicit curricular content and of selected pedagogical strategies on moral reasoning development. Using causal modelling to control for a range of student background variables as well as Time 1 scores, 52% of the variance in moral reasoning scores was explained; we found that these scores were affected by type of curricular content and by three pedagogical strategies (active learning, reflection and faculty-student interaction). Students who experienced more negative interactions with diverse peers were the least likely to show positive change in moral reasoning as a result of participating in any course. Implications for the design of intervention studies are discussed, including the need to attend to selection and attenuation effects. (Contains 1 figure, 4 tables, and 2 notes.)