Background The main toxicity of irinotecan in advanced colorectal cancer (CRC) is delayed diarrhoea. Intestinal SN-38, released by deconjugation of the parent glucuronide excreted into the bile or produced in situ by intestinal carboxylesterase, is toxic to the intestinal epithelium. The canalicular transport of irinotecan and SN-38G is mediated by ABCC2 (MRP2) and ABCB1 (MDR1) which are both inhibited by ciclosporin. We tested whether irinotecan and ciclosporin was non-inferior for anti-cancer efficacy and superior for toxicity compared with single-agent irinotecan. Methods Six hundred and seventy-two patients with advanced, measurable CRC following prior fluoropyrimidine- containing chemotherapy were randomised to either irinotecan 3-weekly 350 mg/m2 (or 300 mg/m2 if age >70 or performance status (PS) = 2) or 3-weekly irinotecan at 140 mg/m2 (120 mg/m2 if age >70 or PS = 2) with ciclosporin 3 mg/kg t.d.s. for three days by mouth starting on the morning before irinotecan. The primary end-point was the proportion of patients alive and progression-free at 12 weeks. The key secondary end-point was the incidence of grade ≥3 diarrhoea within 12 weeks of randomisation. Results The proportion of patients progression-free at 12 weeks with irinotecan was 53.4% compared to 47.2% with irinotecan plus ciclosporin (difference = -6.3%, 95% confidence interval (CI) [-13.8%, 1.3%]). Since the lower limit of the 95% CI crossed the pre-specified non-inferiority margin of -10.6%, non-inferiority of irinotecan plus ciclosporin compared to irinotecan alone was not statistically demonstrated. 15.0% patients developed severe diarrhoea on irinotecan compared to 13.8% on irinotecan plus ciclosporin, a non-significant difference. Interpretation The pharmacokinetic biomodulation of irinotecan using oral ciclosporin does not improve the therapeutic index of irinotecan in advanced CRC. Funding The trial was funded by Cancer Research UK and supported by Amgen Pharma. © 2013 Elsevier Ltd. All rights reserved.