The authors present results from a comprehensive field experiment at Swiss high schools in which they compare the effectiveness of teaching methods in economics. They randomly assigned classes into an experimental and a conventional teaching group, or a control group that received no specific instruction. Both teaching treatments improve economic understanding considerably, while effect sizes between teaching treatments are almost identical. However, preexisting economic competencies crucially affect learning outcomes as more competent students seem to benefit disproportionately from classroom experiments, while weaker students lose out. Supplemental data indicate that the experimental treatment crowded out time for adequately discussing the subject, which may have limited less competent students to generate a profound understanding.