The authors reviewed 42 empirical studies of teaching of critical thinking skills in postsecondary education published between 1994 and 2009. The instructional intervention, test measure, and research design of the studies were analyzed. Study results suggest that: (1) the same instructional interventions can lead to different results, depending on the intervention's implementation; (2) qualitative data can inform researchers about intervention effects that are not easily captured by quantitative instruments; and (3) most studies reviewed are subject to limitations in research design, sample size, or sample representativeness. The following recommendations are made: (1) statistical significance should not be the only criterion for instructors to consider when choosing new teaching methods; (2) multiple test measures, including quantitative and qualitative, should be used to assess changes in students' critical thinking skills; (3) future research should properly address internal validity threats, e.g. by adopting at least a quasi-experimental design, in order to establish causal relationship between intervention and changes in students' critical thinking skills. (Contains 5 figures.)