Imagery training is a well-known technique in sport psychology that it is often applied to improving skill performance in a range of sports. In this study, the central aim was to experimentally examine the effect of different imagery frequencies (3, 4, 5 imagery sessions per week) on basketball shooting performance. We applied a new imagery dose-response protocol, in which we varied frequency, but systematically held the other two key imagery dose variables (repetitions and duration of sessions) constant. Participants were 40 male basketball players (Mage = 20.92, SD = 3.01) who were allocated into four conditions: 3 imagery sessions per week, 4 imagery sessions per week, 5 imagery sessions per week, and a control condition. All 3 imagery conditions had 4 weeks of imagery training. For all four conditions, we measured free throw shooting (FTS) at pre-test, Week 1, 2, 3, post-test, and retention test (Week 5). Control condition participants performed their usual basketball practice with no imagery training. Results showed that the 4 imagery sessions per week condition had the highest FTS means at post-test and retention test, with FTS means that were significantly higher than the control condition at post-test and retention test. The findings and information form this study could contribute to the design of effective imagery training by supporting athletes and coaches to tailor imagery programs. Moreover, the imagery dose-response protocol utilised in this study has potential application to further examine imagery dose-response relationships.