Aim: The aim of this study was to assess the effects of three different training protocols with 8 months in functional fitness and body composition of elderly women. Material and methods: Eighty community-dwelling elderly women were randomly divided into four groups: (i) control group (68.55 ± 5.24 years); (ii) multicomponent training group (68.70 ± 6.51 years), (iii) muscle power training group (66.30 ± 3.87 years) and finally, (iv) resistance training group (67.10 ± 5.46 years). A total body scan in dxa equipment assessed the body composition. The functional fitness (ff) was evaluated with the functional fitness test. A multivariate analysis (anova) with repeated measures on each dependent variable allowed to detect differences in each group and time. The bonferroni test allowed to assess the group x time interaction. Anova revealed no significant effect on time x group interception in body composition variables. Results: A significant effect on intercept time x group in tests of upper body flexibility in control group (p=0.028), multicomponent training group (p=0.001) and muscle power training group (p=0,002) and lower body flexibility multicomponent training group (p<0.001), muscle power training group (p=0.003); and resistance training group (p=0.002); aerobic endurance in multicomponent training group (p<0.001); upper body strength in multicomponent training group (p=0.005), muscle power training group (p=0.008) and resistance training group (p=0.007) and lower body strength in multicomponent training group (p=0.045) and muscle power training group (p=0.006) were founded. Discussion: Our study suggests that, independently of type of training, independent community-dwelling elderly women should improve their functional performance with appropriate exercise programs. Conclusion: We can conclude that 8 months training programs did not induce significant changes to body composition. However, changes in ff were observed.