Summary The aim of the study was to investigate the therapeutic value of reinduction therapy with the same cytostatic treatment that had been used for induction treatment in patients with metastatic colorectal cancer. A total of 71 patients, all of whom had responded or achieved stable disease lasting >12 weeks after six monthly courses of first-line treatment with 5-fluorouracil + racemic leucovorin (5-FU/LV; n = 35) or 5-FU plus the I-isomer of LV (LLV; n = 34) were entered in this study. At the time of relapse, the same treatment was used for initial therapy: racemic LV or LLV was administered at 100 mg m-2 day-' by i.v. bolus injection, immediately followed by 5-FU 400 mg m-2 day-1 given as a 2-h infusion. Chemotherapeutic drugs were given on 5 consecutive days at 4-week intervals x 6 or until there was evidence of tumour progression. Among 49 evaluable patients, reinduction therapy that was initiated after a median treatment-free interval of 5.4 months (range 3-14.5) resulted in nine partial response (PR) (18%) and 26 stable disease (SD) (53%), yielding an overall tumour control rate of 69% (95% confidence interval, 54.6-81.7%). The median time to treatment failure from reinduction was 6.4 months, and the median survival duration from reinduction was 8.9 months (20.1 months as judged from the beginning of induction therapy). The toxicity associated with retreatment was generally mild to moderate; compared with initial treatment, there was no significant difference in terms of the overall rate (P = 0.33) or severity (P = 0.19) of adverse reactions. Our data suggest that in patients with advanced colorectal cancer an interrupted treatment strategy, i.e. retreatment with the same regimen in case of relapse .3 months after discontinuation of 6 months of successful treatment with 5-FU/LV or 5-FU/LLV is an acceptable therapeutic concept.Keywords: advanced colorectal cancer; chemotherapy; reinduction; 5-fluorouracil; leucovorin Advanced colorectal cancer remains a therapeutic challenge to clinicians involved in the management of this common malignant disease, which continues to be the second leading cause of cancer death in the Western world (Chu et al, 1994). Despite intensive efforts to improve the poor prognosis of patients with advanced colorectal cancer, therapeutic progress has been hampered by its apparent chemotherapeutic refractoriness. Although randomized trials have established with reasonable certainty that fluorouracil/leucovorin-based chemotherapy results in substantive therapeutic gain compared with best supportive care (The Nordic Gastrointestinal Tumour Adjuvant Therapy Group, 1992;Scheithauer et al, 1993) or treatment with 5-FU alone (Poon et al, 1989; Advanced Colorectal Cancer Meta-Analysis Project, 1992), there is considerable room for further improvement in terms of the response rates and tolerance of treatment. Until now, it has not been possible to define the optimal regimen of biochemical modulation of fluoropyrimidines (Doroshow, 1996), and there is also sparse information in terms of the...