Conducting classroom-based educational research trials is important for establishing the efficacy and effectiveness of specific instructional interventions. Such endeavours, however, are challenging to implement. This was made evident during a recent independent evaluation of the efficacy of the MiniLit program, wherein various difficulties emerged relating to the dosage and fidelity of instruction, and the measures and analyses employed by the research team. As such, this served as an object lesson in what can, and frequently does, go wrong in even the best planned intervention research enterprises conducted in schools. The present article is intended to capture the authors’ experiences in implementing research trials in school contexts, with specific examples drawn from the independent evaluation of MiniLit. In particular, this study has reinforced the need to select assessment measures carefully, according to how well they represent targeted skills in the specific population of interest. In addition, it has highlighted the importance of planning program efficacy trials such that participants can receive enough exposure of the intervention to progress to a realistic extent.