This study assessed the consequences of the reduced use of Artificial Insemination (AI) on genetic gain with constraints on inbreeding changes for a repeated maternal trait in meat sheep. Using a stochastic model, scenarios inspired by the French context were compared and defined by different strategies (no AI, AI without and with a progeny test of young rams), rate of AI (from 0 to 80%) and proportions of pedigreed females born from Natural Mating (NM). Additional gains from +15 to + 50% were obtained for the AI-based design without progeny testing, and from + 57 to + 84% for the AI-based design with progeny testing compared to the no AI design. When only part of the females were from a known sire, genetic gain strongly decreased in no AI breeding programs, with a reduced effect when the rate of AI increased.