Background: The aim of this study was to investigate the effect of exercise during pregnancy on early childhood neurodevelopment (cognitive, motor, and language domains). Methods: A randomized controlled trial nested into the 2015 Pelotas (Brazil) Birth Cohort was conducted. Healthy pregnant women were enrolled between 16 and 20 weeks of gestation; 424 women and their children (intervention [n = 141]; control [n = 283]) were analyzed. An exercise-based intervention 3 times per week was delivered over 16 weeks. Child neurodevelopment and its domains were assessed at 1, 2, and 4 years. Standardized mean differences and 95% confidence intervals are presented. Results: No effects of exercise during pregnancy on child neurodevelopment and its domains at age 1 year were observed. Compared with the control group, children from women in the exercise group had higher language score at age 2 years (standardized mean differences = 0.23; 95% confidence intervals, 0.02 to 0.44) and higher cognitive score (standardized mean differences = 0.22; 95% confidence intervals, 0.03 to 0.41) at age 4 years. No effects of exercise during pregnancy were observed in the motor domain at 1, 2, and 4 years. Conclusions: No detrimental effects of exercise during pregnancy on child neurodevelopment were observed. In addition, these findings suggest that exercise during pregnancy can result in small benefits for language and cognitive development.