Background: The Rotterdam Evaluation Study of Stuttering Therapy randomized trial (RESTART) was seminal, comparing the Lidcombe Program with RESTART Demands and Capacities Model-based treatment (RESTART-DCM) for pre-school age children who stutter. Aims: To critique the methods of the RESTART trial to develop guidelines for its systematic replication and extension. Beyond that, to contribute to the refinement of existing methodological guidelines for early stuttering intervention. Method: The discussion is organized around methodological issues of primary outcomes, treatment completion, clinician allegiance, treatment fidelity, age exclusions and no-treatment control reasoning. Main Contribution: We recommend six methodological guidelines to guide future clinical trials comparing the Lidcombe Program with RESTART-DCM, which can be applied to clinical trials of other early stuttering intervention methods: (1) incorporate a continuous measure of primary outcome; (2) ensure that all children in clinical trials have completed treatment; (3) eliminate potential bias due to clinician allegiance; (4) establish treatment fidelity within and beyond the clinic; (5) include children younger than 3 years in clinical trials; and (6) establish an estimate of treatment effect size at some stage of treatment development. Conclusion: In addition to guiding future clinical research comparing RESTART-DCM and Lidcombe Program treatment, these recommendations may extend to influence positively other treatment developments for early stuttering.