Objective: The objective of this study was to assess supervised exercise training on gestational weight gain (GWG) and body fat gain (BFG) in late pregnancy. Design: A randomized controlled trial. Setting: East Carolina University. Sample: 3 exercise types (aerobic n = 66, combination n = 48, or resistance n = 51) and control group (n = 54). Methods: Exercise was conducted 3 times/wk from 16 weeks gestation until delivery. Main outcome measures: GWG, BFG, and skinfolds measurements at 16 and 36 gestational weeks. Analyses were performed using both intention-to-treat (ITT) and per-protocol (>75% adherence). Results: Per-protocol results show differences between groups for 16- to 36-week change in maternal skinfold sum (0.02), body fat % (0.01) and fat free mass % (0.02). Per-protocol results stratified for pre-pregnancy BMI showed differences in healthy weight participants for changes in maternal skinfold sum (0.01), body fat % (0.004) and fat free mass % (0.01), and in overweight participants, from 16 to 36 weeks gestation, we noted changes in skinfold sum (0.03), body fat (0.02) and fat free mass (0.03). No differences were found for GWG between exercise groups for both ITT and per-protocol analysis similarly, there were no differences between healthy weight and overweight participants for per-protocol analysis. Type of exercise (0.01) was a significant predictor of change in fat free mass percentage (0.04) and BFG percentage (R ² 0.03) from 16 to 36 weeks gestation. Conclusions: Exercise during pregnancy effectively reduces maternal body fat gain.