The purpose of this study was to conduct a preference trial as a preliminary test of preference effects on teacher behavior relative to implementation (adoption, adherence, quality). Teachers were randomly assigned to "preference" or "no-preference" groups and then trained to implement the intervention. Direct observation occurred immediately after initial training, after 6 weeks of coaching support, and after 4 weeks of no support. Results showed that, when compared with the no-preference group, teachers who had the opportunity to exert a preference adopted the intervention sooner and sustained higher fidelity and quality of implementation independent of coaching. Furthermore, though most teachers in the no-preference group did adopt the intervention and demonstrate high fidelity following coaching, implementation did not sustain after the withdrawal of coaching.