Caffeine and dextromethorphan have been used successfully both alone and in combination to assess phenotype and enzyme activity in children of various ages. Previous pediatric phenotyping studies with these agents have used varying durations of urine collection. However, the minimum duration required for accurate phenotypic assessment with these compounds in children remains unknown. We calculated the cumulative metabolite recoveries and molar ratios in urine collected from children for 2, 4, 6, and 8 hours after caffeine and dextromethorphan administration to determine when respective urinary molar ratios stabilize and thus likely accurately reflect enzyme activity. Subjects (n = 24, ages 3-8 years) were given 4 oz of Coca-Cola(R) ( approximately 11.5 mg caffeine) and a single oral dose of dextromethorphan (0.5 mg/kg). Urine was collected at discrete intervals (0-2, 2-4, 4-6, and 6-8 h) during an 8-hour period, and the cumulative metabolite recoveries and urinary molar ratios were calculated. CYP2D6 genotyping was also performed in 21 of 24 subjects. In CYP2D6 extensive metabolizers, the extent of recovery for relevant metabolites was equivalent by 4 hours and represented 45% to 60% of the total amount recovered in the 8-hour period. The 2-hour CYP1A2 ratio was significantly different from those of longer collection intervals. Metabolite ratios for all other enzymes (i.e., NAT-2, XO, and CYP2D6) were independent of the duration of urine collection. These data suggest that a 4-hour urine collection is adequate for the concurrent assessment of hepatic CYP1A2, NAT-2, XO, and CYP2D6 activity in children ages 3 to 8 years who are CYP2D6 extensive metabolizers, using standard caffeine and dextromethorphan phenotyping methods. Longer collection periods may be required, however, in younger children or CYP2D6 poor metabolizers.