BACKGROUND: Implementation science (IS) trials in HIV treatment and prevention evaluate implementation strategies that deliver health-enhancing tools such as antiretroviral medicines or prevention technologies to those who need them, rather than evaluating the tools themselves. METHOD: Opinion piece drawing on a non-systematic review of HIV prevention and treatment trials to inform an assessment of 5 key challenges for IS trials. RESULTS: Randomized controlled trials (RCTs) are an appropriate design for IS but must address 5 challenges. IS trials must be feasible to deliver, which will require addressing challenges in maintaining multisectoral partnerships, strengthening routine data, and clarifying ethical principles. IS trials should be informative, evaluating implementation strategies that are well designed and adequately described, and measuring implementation outcomes, coverage of tools, and, when appropriate, epidemiological impacts. IS trials should be rigorous, striving for internally valid estimates of effect by adopting best practices, and deploying optimal nonrandomized designs where randomization is not feasible. IS trials should be relevant, considering and documenting how "real-life" is the implementation monitoring and whether research participants are representative of the target population. Finally, IS trials should be useful, deploying process evaluations to provide results that can be used in onward decision-making. CONCLUSIONS: IS trials can help ensure that efficacious tools for HIV prevention and treatment have maximum impact in the real world. These trials will be an important component of this scientific agenda if they are feasible to deliver and if their results are informative, rigorous, relevant, and useful.