Capecitabine is a fluoropyrimidine-based drug that offers physicians a more convenient treatment for advanced colorectal cancer (CRC), with manageable toxicity and antitumor activity comparable to that of continuous-infusion therapies with 5-fluorouracil (5-FU). However, there are no validated and established predictive factors for clinical outcome of capecitabine efficacy in CRC. The gene expressions of the pyrimidine metabolism enzymes dihydropyrimidine dehydrogenase (DPD), thymidine phosphorylase (TP) and thymidylate synthase (TS) have previously been shown to be response determinants of fluoropyrimidine-based drugs in various tumors. Therefore, we investigated whether intratumoral mRNA expression levels of these genes are also associated with the clinical outcome of patients with metastatic CRC treated with first-line capecitabine. Thirty-seven patients with metastatic CRC were enrolled in this study and treated with single agent capecitabine. The intratumoral mRNA levels of DPD, TP and TS were assessed from paraffin-embedded tissue samples using laser-capture-microdissection methods and quantitative real-time PCR. There were 20 women and 17 men with a median age of 61 years (range 49-74). The median progression-free survival was 6.7 months (95% CI, 4.8-11.6 months), with a median follow-up of 14.4 months (range 1.3-18.7 months). Complete response was observed in 1 (3%), partial response in 6 (20%), stable disease in 14 (47%) and progressive disease in 9 (30%) patients (response was inevaluable in 7 patients). Higher gene expression levels of DPD were associated with resistance to capecitabine (P=0.032; Kruskal-Wallis test). Patients with a lower mRNA amount of DPD (<or=0.46) had a longer progression-free survival compared with patients that had a higher mRNA amount (8.0 vs. 3.3 months; adjusted P=0.048; log-rank test). This pilot study suggests that intratumoral gene expression levels of DPD may be useful in predicting the clinical outcome of patients with metastatic CRC with first-line single agent capecitabine treatment. Our data should be validated in larger and prospective clinical trials.