Bioelectrical impedance analysis (BIA) has been shown to be highly related to skeletal muscle mass and blood volume, both of which are important determinants of maximal oxygen uptake (VO(2max)). The aim of the present study was therefore to investigate the ability of whole-body and segmental multi-frequency BIA to improve current nonexercise VO(2max) prediction models. Data for VO(2max) (mL min(-1)), anthropometry, self-reported physical activity (PA-R) and BIA were collected in 115 men and women. Multiple linear regression analysis (MLR) was used to develop the most parsimonious prediction model. Segmental BIA was not superior to whole-body measurements. Correlation coefficients between VO(2max) and resistance indices were significantly higher at 500 kHz compared to 50 kHz (p < 0.05). Intracellular resistance index, however, showed the highest correlation with VO(2max) (r = 0.89). After adjusting for age, gender and PA-R, MLR revealed that the inclusion of intracellular resistance index was slightly, but significantly (p < 0.001), superior to models based on anthropometry. Subgroup analyses indicated that the true benefit of BIA might be most prevalent in subjects with particularly low VO(2max) (<2500 mL min(-1)). In short, whole-body BIA marginally improves the accuracy of nonexercise VO(2max) prediction models and its advantage is most pronounced in individuals with particularly low VO(2max).