Despite growing knowledge of pregnancy-induced changes in physiology that may alter maternal and fetal pharmacokinetics, and therefore drug efficacy and safety, evidence-based antenatal doses are lacking for most drugs. Pharmacokinetic models and expanding clinical data in pregnancy may support antenatal doses. In this article, we introduce a comprehensive and user-driven Framework for Dose Selection in Pregnancy (FDSP), developed and validated to support the clinical implementation of best-evidence and in some cases, model-informed doses for pregnant women and/or fetuses. After initial development and validation by experts, the framework prototype was piloted to formulate an antenatal dosing strategy for sertraline in depression and anxiety disorders. Next, the framework was validated and assessed for usability by a multidisciplinary working committee of end-users comprising healthcare practitioners, experts from other disciplines including pharmacometrics, reproductive toxicology and medical ethics, alongside pregnant women and a partner. The resulting framework encompasses the following: rationale for drug selection, a comprehensive analysis of pharmacokinetic and dose-related efficacy and safety data, and implementation aspects including feasibility and desirability of the recommended antenatal dose based on a structured maternal and fetal benefit-risk assessment. An antenatal dose recommendation for sertraline, as a proof-of-concept, was formulated using this approach and endorsed for clinical use by the working committee. The FDSP, as demonstrated by the example of sertraline, is fit for supporting the development of best-evidence acceptable and clinically feasible antenatal doses.