In work-related instrumental learning contexts, the role of reflective activities is unclear. Kolb's experiential learning theory and Mezirow's transformative learning theory predict skill adaptation as an outcome. This prediction was tested by manipulating reflective activities and assessing participants' response and error rates during novel applications of an instrumentally learned skill (skill adaptation). Participants were randomly assigned to three conditions (interference, reflection, or critical reflection) according to blocking variables: gender, age, and reflective propensity. Participants then completed behavioral skills training with embedded reflective activities. Afterward, participants completed a novel application task. Analyses of variance did not reveal differences in response or error rates among conditions or an interaction between reflective activity and reflective propensity on response rate. An interaction was found between reflective activity and reflective propensity on error rate: In the critical reflection condition, nonreflective learners had higher error rates than reflective learners. Potential explanations and implications for research and practice are offered.