In 38 females, aged 21.3 ± 1.8 years (mean ± SD) and 22 males, aged 21.6 ± 2.0 years, total body water (TBW) was assessed by bioelectrical impedance at a frequency of 100 kHz (Z100), and extracellular water (ECW) was assessed at a frequency of 1 (Z1) or 5 kHz (Z5). Impedance ratios of low frequencies (Z1 or Z5) to high frequency (Z100) were calculated as indicators of body water distribution. Furthermore, changes in body water distribution during the menstrual cycle were assessed in 16 females, aged 22.1 ± 1.7 years, divided into subjects using or not using oral contraceptives. In general, the ratio ECW to TBW was lower in males compared to females, which was reflected in the ratios Z1/Z100 and Z5/Z100·However, the ratio Z5/Z100 reflected this difference less clear, probably because at 5 kHz the current partly passes the cell membranes. Weight changes during the menstrual cycle are generally accepted to be due to a retention of ECW. The ratio Z5/Z100 and the ratios Z1/Z100and Z5/Z100 for those subjects not using oral contraceptives differed significantly between day with minimal and day with maximal weight. A negative correlation was found between weight change, still considering day with minimal and day with maximal weight, and change in impedance at 5 and 50 kHz for subjects not using oral contraceptives. It is concluded that differences in body water distribution are reflected by low-to-high impedance ra