To investigate the effects of resistance training (RT) program on body composition, functional capacity, inflammatory and oxidative stress biomarkers in sarcopenic obesity elderly women. This study is a randomized controlled trial. Thirty-seven sarcopenic obesity (SO) elderly women were randomly allocated into two groups: a training group (TG, n=18) or a non-exercising control group (CG, n=19). A supervised RT program was performed by TG for 12 weeks and CG did not perform any type of physical exercise during this period. Body composition (DXA), functional capacity [(muscular strength, and the 10 m walk test (10MW), and rising from sitting position test (RSP)] and blood sample measurements (after a 12 h fasting) were performed pre- and post-training. The investigation was carried out over a period of 16 weeks, with 12 weeks dedicated to the RT program and 4 weeks allocated for measurements. Anthropometric, muscular strength, body composition, and blood samples measurements were performed at weeks 1-2, and 15-16. A supervised progressive RT was performed between weeks 3-14. The CG did not perform any type of physical exercise during this period. However, given the possible positive effects of exercise, elderly women of this group were incorporated into a new exercise program after the end of this experiment. Data distribution was tested using the Shapiro Wilk test. Descriptive statistics are presented as means and standard deviations. Student's independent t-test and chi-square test were used to compare the control and intervention groups regarding the general characteristics and clinical conditions (categorical variables), respectively. Two-way analysis of covariance (ANCOVA) for repeated measures was applied for comparisons, with baseline scores used as covariates. When F-ratio was significant, Bonferroni's post hoc test was employed to identify the mean differences. Between group effect size (ES) was calculated to verify magnitude of the difference between groups. The Z-score of the percentage changes (from pre- to post-training) of the raw data for each parameter was calculated, and a total Z-score, derived from the all components was calculated. To verify the differences among groups on total Z-scores, a test T independent also was applied. For all statistical analyses, significance was accepted at P< 0.05. The data were analyzed using STATISTICA software version 13.2 (Statsoft Inc., Tulsa, OK, USA).