Measuring the bioelectrical impedance (BI) is a simple and non-invasive method for estimating body fat or muscle mass. However, body impedance is affected by variations in the distribution of body fluid without reference to actual body fat or muscle mass. Twenty healthy college students (10 males, 10 females; mean age 21.0±2.3 years) participated in the study. Their mean body mass index was 20.7±2.6 kg/m 2 . Bipolar electrodes were place on all extremities, and InBody 3.0 TM (Biospace Co., Ltd., Seoul, Korea) was used to measure bioelectrical impedance. Each subject remained in a supine position on a comfortable bed between 07:00 and 12:00 except for excretion and measurement of BI. BI was measured hourly using frequencies ranging from 5 to 500 kHz. The subjects refrained from eating, drinking and exercising between 07:00 and 12:00 during the first week of measurements, and drank 6.7 ml/kg of water at 07:00 after the first measurement of BI during a subsequent week of measurements. Bioelectrical impedance was higher in female subjects in all body segments and conditions (p<0.01). BI in the right arm was lower than that in the left in all participants (p<0.001). The difference between the highest and lowest BI among six measurements was largest in the upper extremities, followed by the lower extremities. Differences in the coefficient of variation CV values of the right arm of both females and males at 50, 250, and 500 kHz during fasting were significantly smaller than after drinking water. Hydration had no effect on the differences in the CV values of the body trunk and lower extremity BI or BI at lower frequencies. BI indicates the possibility of remarkable decrease in variation in the upper extremity BI at higher frequencies by taking 6.7 ml/kg of water at get up and enables minimizing the estimate error of body fat percentage.