The aim of the present study was to examine (a) the relationship of body fat (BF) assessed by bioimpedance analysis (BIA) and skinfold thickness (SKF) and (b) the variation of BF by age depending on the assessment method. Participants were 32 women and 134 men recreational marathon runners, who were tested for BF using both assessment methods (BIA and SKF). Rc between BIA and SKF assessment methods was 0.803 (95% CI; 0.640, 0.897) in women and 0.568 (95% CI; 0.481, 0.644) in men. A large main effect of the assessment method on BF was observed ( < 0.001, = 0.156) with SKF presenting higher BF than BIA by 2.9%. The difference between SKF and BIA was 3.9 ± 2.7% (95% confidence intervals, CI; 3.4; 4.3, < 0.001) in men, whereas no difference was found in women (-0.9 ± 2.9%; 95% CI; -1.9; -0.2, = 0.101). BF correlated with age with small magnitude (BIA, = 0.18, = 0.036; SKF, = 0.23, = 0.007) in men, i.e., the older the age, the higher the BF. A similar trend of moderate magnitude was observed in women for BIA ( = 0.45, = 0.011), but not for SKF ( = 0.33, = 0.067). In conclusion, practitioners involved in the training of recreational runners would be advised to consider that BIA elicits a lower BF value than the SKF method in men.