The purpose of this study was to verify the effects of 16 weeks of practicing different exercise programmes on body composition. This is an exploratory and descriptive study of 89 women aged 25 to 55 years (41.42 ± 9.23 years). The subjects were randomly divided into three experimental groups (EG): practitioners of strength training (SG), dance (DG), hydrogymnastics (HG), and a control group (CG) with sedentary women. Measurements of body mass and height, circumferences of the chest, waist, abdomen, hips, thighs, calves, and skinfolds of the triceps, suprailiac and thigh were registered in three different moments: prior to the commencement of the training program, again after 8 weeks of training, and finally after 16 weeks of training. Body density was estimated by using the trifold protocol by Jackson, Pollock and Ward. The ANOVA and deltas of change (Δ%) were used for data analysis. The level of significance was set at p<0.05. The effects of greater statistical significance on body composition related the variables "time", "group" and the interaction between the two (time x group) were observed for the percentage of fat - F% (F (1.79, 152.52) = 24.59, p <0.001, η 2 = 0.22), fat mass - FM (F (1.75, 149.01) = 12.65, p <0.001, η 2 = 0.13) and lean mass - LM (F (1.77, 150.66) = 47.38, p <0.001, η 2 = 0.36). The HG and SG were more beneficial in reducing F%. It was observed that the EG indicated healthier anthropometric aspects compared to the CG, regardless of the type of exercise programmes practiced. The time factor was more representative over the effects of exercise on anthropometric dimensions.