The present investigation was performed to confirm the relationship between the circadian variation of microproteinuria and physical activity. Urine samples from 10 normal male volunteers, collected during six consecutive 4-h periods, were examined for albumin, alpha 1-, beta 2-microglobulin, NAG, electrolytes and hormones. The fluctuations in heart rate (HR) and blood pressure (BP) over 24-h were measured at 30-min and 1-h intervals, respectively. Energy expenditure (EE) was calculated using the equation of regression between HR and oxygen uptake measured on another day. The variations of HR (delta HR) and EE (delta EE) based on a 24-h average (bpm and kcal/kg/h) were used as indices of change in physical activity during an ordinary day. The correlation coefficients between delta HR and the variations of albumin (delta Alb) and beta 2-microglobulin (delta beta 2M) from the 24-h average (micrograms/h.cr 1 mg) were 0.619 and 0.670 (p less than 0.001), respectively. Increased excretions of both glomerular and tubular proteins were correlated with the increase in HR and/or EE during daytime activity. During rest time at night, the variations in alpha 1M, beta 2M and NAG excretion were different from the variations in albumin. A temporary inhibition of tubular protein excretion was observed only in the early morning (04:00-08:00), although albumin excretion was inhibited throughout the nighttime. These findings suggested that physical activity may influence the diurnal variations in protein excretions, that albuminuria may be more sensitive to daytime activity, and that fluctuation of tubular protein excretion may be preferably controlled by an endogenous mechanism. Timed overnight or first-morning urine may be recommendable as a sample for determination of microalbuminuria for screening of clinical diabetic nephropathy.