BACKGROUND In Nordic countries, the standard treatment of colorectal cancer (CRC) in the adjuvant setting is bolus 5-fluorouracil (5-FU) plus leucovorin alone or in combination with oxaliplatin. 5-FU competes with the natural occurring pyrimidine uracil (Ura) as a substrate for dihydropyrimidine dehydrogenase (DPD; enzyme commission number 1.3.1.2). Low DPD activity is associated with toxicity during treatment. Pretherapeutic detection of DPD deficiency could prevent severe toxicity otherwise limiting drug administration. Assays showing that DPD deficiency impairs breakdown of Ura to dihydrouracil (UH(2)) seem promising for clinical use. METHODS Urine was collected from 56 untreated volunteers and 143 patients with CRC before adjuvant treatment. Ura and UH(2) were analyzed using a column-switching high-performance liquid chromatography method that incorporates reversed-phase and cation-exchange columns. Ura, UH(2), and UH(2)/Ura levels were related to toxicity. RESULTS Ura and UH(2) in patients were not different from controls. UH(2) was significantly higher in women compared with men. The UH(2)/Ura ratio, however, did not differ according to sex. Low UH(2) and UH(2)/Ura levels were associated with diarrhea in men. Women experiencing thrombocytopenia had significantly higher Ura compared with women with no thrombocytopenia. The UH(2)/Ura ratio correlated negatively with total toxicity score in men (r = -0.39, P = .020). CONCLUSIONS Pretherapeutic Ura and UH(2) levels per se may be related to risk of side effects during adjuvant 5-FU-based treatment, whereas the UH(2)/Ura ratio may not always reveal such a risk. Sex is a strong risk factor for toxicity, showing the importance of evaluating male and female patients separately.