Studies on short-term renal responses to (pharmacological) intervention require accurate and multiple collection of urine samples. Several invasive techniques have been described for frequent urine collection of the conscious rat, each having their own limitations. No data are available about the feasibility of the spontaneously voiding, freely moving rat for this purpose. In the present study, bladder voidings of six rats were time-registered and collected separately for several days. The data show a considerable 24-h variation coefficient of both the voided volume and the bladder collection time with a poor correlation between the two parameters. Forced diuresis induced by continuous i.v. infusion (2 ml/h) increased the frequency of urine voiding and thus the time-resolution of the urine-production pattern. However, this method failed to reduce the variation coefficient of the voided volume, the collection time, and the correlation between the two parameters. The fact that variations in creatinine excretion paralleled the variation in urinary flow suggests that both phenomena are likely be due to incomplete bladder emptying. Correction for this incomplete bladder collection, using the creatinine excretion, indeed reduced the variation coefficient of sodium excretion successfully from 61 +/- 17% to 29 +/- 5% during normal diuresis and from 56 +/- 19% to 22 +/- 6% during forced diuresis. In conclusion, the spontaneously voiding, freely moving rat can be used for short-term renal response studies if the collected urine samples are corrected for incomplete bladder emptying using urinary creatinine concentrations. This procedure allows the detection of changes in a urinary parameter if this exceeds a 40% deviation of the normal value.