This study examines the effectiveness of teaching pragmatic information through the use of explicit or implicit pre-instruction, and explicit or implicit feedback, to English-speaking learners of third-semester Spanish. Results on a pretest, posttest, and delayed posttest indicate that the groups that experienced explicit pre-instruction and explicit feedback during exercises performed significantly better than the other experimental group and the control group in multiple choice items. The group that had implicit instruction together with implicit feedback performed significantly better in the open-ended dialogues, suggesting that the question recasts used had a positive effect on their pragmatic production. The delayed posttest, however, revealed that such gains are not clearly retained in the longer term. Nevertheless, the two posttests indicate that those groups that received instruction and feedback, whether explicit or implicit, appear to become aware of a greater number of options to express suggestions, and also of a need for pragmatic mitigation, more quickly than the control group. These findings are encouraging for the use of pragmatic instruction in the classroom to develop a greater pragmatic competence. (Contains 2 figures and 5 tables.)