•Changes in tube potential in CT acquisition changes muscle attenuation values.•Changes in tube potential changes skeletal muscle index and steatotic muscle area.•CT acquisition parameters should be standardized for body segmentation. Our purpose was to investigate whether tube potential in contrast-enhanced computed tomography (CT) affects body composition analysis. Images from dual-source, dual-energy CT from the abdomen with intravenous contrast media administration were used. A total of 17 patients (11 women, mean age 52) with a mean body mass index of 20.8 kg/cm2 were included. Simultaneously acquired images with a tube voltage of 80 kV and 140 kV were compared. Body composition was analyzed on a single slice at the L3 level. Parameters evaluated included muscle and fat attenuation (Hounsfield units HU), skeletal muscle index (cm2/m2), muscle area (cm2), and steatotic muscle area (cm2). Significant differences between 80 kV and 140 kV series were compared using the paired Student's t test. Tube potential affected muscle attenuation with an average difference of 17% between 80 kV and 140 kV series (48 HU versus 41 HU, P < 0.01), fat attenuation (−84 HU versus −69 HU, P < 0.01), skeletal muscle index of 5.2% (40.1 cm2/m2 versus 42.2 cm2/m2, P < 0.01), muscle area of 5.1% (117 cm2 versus 123 cm2, P < 0.01), and steatotic muscle area of 12.9% (31 cm2 versus 35 cm2, P < 0.01). Tube potential significantly affects body segmentation in contrast-enhanced CT.