This quasi-experimental study adopted a pretest/posttest design to investigate the effect of instructional intervention in teaching polite refusal strategies explicitly on Iranian EFL learners' performance of the speech act of refusing. The participants, consisting of 24 male elementary EFL learners aged 12-18, responded to a discourse completion task (DCT) prior to and after they had been provided with explicit instruction concerning the polite performance of refusals in English. Adopting Brown and Levinson's (1987) politeness theory and Beebe, Takahashi and Uliss-Weltz's (1990) taxonomy of refusal strategies, the researchers found that the participants' refusal semantic formulas in the pretest contained a variety of impoliteness markers including directness, lack of mitigation, and terseness of responses. The pedagogical instruction was directed at eliminating these inappropriacy elements, addressing the lengthening and intensification of refusal semantic formulas, use of adjuncts to refusals, titles, honorifics, apologizing, etc. The participants' responses to the DCT in the posttest showed a high level of appropriacy in the semantic content of refusal utterances compared to their responses in the pretest. Furthermore, the findings demonstrated that significant differences existed in terms of the content, frequency and types of both refusal strategies and adjuncts to refusals between the pretest and posttest phases. In conclusion, the study revealed the positive effects of instructional intervention on the development of the pragmatic competence of learners with low linguistic proficiency levels.