The relationship between urinary ethanol concentration, urine/blood ratio of ethanol, and urinary creatinine content was investigated by the analysis of two successive voids from 40 individuals apprehended for driving under the influence of alcohol (DUI) in Sweden. The first specimen of urine was collected 24 +/- 17 min (mean plus or minus standard deviation) before sampling blood, and the second specimen was collected 46 +/- 30 min after blood sampling. The mean blood-alcohol concentration (BAC) was 2.21 +/- 0.76 g/L, and the corresponding concentrations in the urine (UAC) were 2.88 +/- 1.08 g/L for the first void and 2.68 +/- 1.01 g/L for the second void. The mean urine/blood ratios of ethanol were 1.31 +/- 0.21 for the first void and 1.20 +/- 0.11 for the second void; the difference of 0.11 +/- 0.24 was statistically significant (t = 3.08, p < 0.01). The concentrations of creatinine in urine from DUI suspects were 0.72 +/- 0.64 g/L for the first void and 0.86 +/- 0.73 g/L for the second void; there was no significant difference (t = 1.45, p > 0.05). The urinary creatinine content in specimens from drunk drivers was significantly less than the first morning voids from 3313 prison inmates (1.78 +/- 0.74 g/L). No significant correlations existed between UAC and urinary creatinine content (r = -0.14) between urine/blood ratios of ethanol and urinary creatinine (r = -0.19). However, the concentrations of ethanol in blood and urine were highly correlated; they were r = 0.94 +/- 0.055 (p < 0.001) for first void and r = 0.96 +/- 0.045 (p < 0.001) for the second void. This study demonstrates that the relative dilution of urine specimens, as reflected in creatinine content, is not associated with the concentration of ethanol in the urine samples or with the UAC/BAC ratio.