The present study aimed to investigate the validity of bioelectrical impedance and ultrasonographic methods for predicting the muscle volume of the upper arm. Bioelectrical impedance of the right upper arm and the thickness of elbow flexor and extensor muscles were measured from 26 healthy young adult men using a specially designed bioelectrical impedance data acquisition system and brightness-mode ultrasound apparatus, respectively. As reference data, the muscle volume (MV(MRI)) of the right upper arm was determined using a magnetic resonance imaging method. The impedance index was calculated as L2/Z, where L is the upper arm length and Z is the impedance of the shoulder to the elbow. The muscle volume (MV(ULT)) was calculated as pi x (d/2)2 L, where d is muscle thickness. L2/Z and MV(ULT) were significantly correlated with MV(MRI) with correlation coefficients of 0.971 and 0.962, respectively. In these relationships, the standard errors of estimation were 44.2 cm3 (6.3%) for L2/Z and 50.7 cm3 (7.2%) for MV(ULT). There was no significant difference between the absolute errors of muscle volumes estimated by L2/Z and MV(ULT): 36.2 (4.8, SEM) cm3 for L2/Z versus 40.3 (5.8) cm3 for MV(ULT). The present results suggest that both bioelectrical impedance and ultrasonographic methods may be useful for predicting the muscle volume of the upper arm.