BACKGROUND: The phe508del CFTR mutation causes cystic fibrosis by limiting the amount of CFTR protein that reaches the epithelial cell surface. We tested combination treatment with lumacaftor, an investigational CFTR corrector that increases trafficking of phe508del CFTR to the cell surface, and ivacaftor, a CFTR potentiator that enhances chloride transport of CFTR on the cell surface. METHODS: In this phase 2 clinical trial, we assessed three successive cohorts, with the results of each cohort informing dose selection for the subsequent cohort. We recruited patients from 24 cystic fibrosis centres in Australia, Belgium, Germany, New Zealand, and the USA. Eligibility criteria were: confirmed diagnosis of cystic fibrosis, age at least 18 years, and a forced expiratory volume in 1 s (FEV1) of 40% or more than predicted. Cohort 1 included phe508del CFTR homozygous patients randomly assigned to either lumacaftor 200 mg once per day for 14 days followed by addition of ivacaftor 150 mg or 250 mg every 12 h for 7 days, or 21 days of placebo. Together, cohorts 2 and 3 included phe508del CFTR homozygous and heterozygous patients, randomly assigned to either 56 days of lumacaftor (cohort 2: 200 mg, 400 mg, or 600 mg once per day, cohort 3: 400 mg every 12 h) with ivacaftor 250 mg every 12 h added after 28 days, or 56 days of placebo. The primary outcomes for all cohorts were change in sweat chloride concentration during the combination treatment period in the intention-to-treat population and safety (laboratory measurements and adverse events). The study is registered with ClinicalTrials.gov, number NCT01225211, and EudraCT, number 2010-020413-90. FINDINGS: Cohort 1 included 64 participants. Cohort 2 and 3 combined contained 96 phe508del CFTR homozygous patients and 28 compound heterozygotes. Treatment with lumacaftor 200 mg once daily and ivacaftor 250 mg every 12 h decreased mean sweat chloride concentration by 9.1 mmol/L (p<0.001) during the combination treatment period in cohort 1. In cohorts 2 and 3, mean sweat chloride concentration did not decrease significantly during combination treatment in any group. Frequency and nature of adverse events were much the same in the treatment and placebo groups during the combination treatment period; the most commonly reported events were respiratory. 12 of 97 participants had chest tightness or dyspnoea during treatment with lumacaftor alone. In pre-planned secondary analyses, a significant decrease in sweat chloride concentration occurred in the treatment groups between day 1 and day 56 (lumacaftor 400 mg once per day group -9.1 mmol/L, p<0.001; lumacaftor 600 mg once per day group -8.9 mmol/L, p<0.001; lumacaftor 400 mg every 12 h group -10.3 mmol/L, p=0.002). These changes were significantly greater than the change in the placebo group. In cohort 2, the lumacaftor 600 mg once per day significantly improved FEV1 from day 1 to 56 (difference compared with placebo group: +5.6 percentage points, p=0.013), primarily during the combination period. In cohort 3, FEV1 did not change significantly across the entire study period compared with placebo (difference +4.2 percentage points, p=0.132), but did during the combination period (difference +7.7 percentage points, p=0·003). Phe508del CFTR heterozygous patients did not have a significant improvement in FEV1. INTERPRETATION: We provide evidence that combination lumacaftor and ivacaftor improves FEV1 for patients with cystic fibrosis who are homozygous for phe508del CFTR, with a modest effect on sweat chloride concentration. These results support the further exploration of combination lumacaftor and ivacaftor as a treatment in this setting. FUNDING: Vertex Pharmaceuticals, Cystic Fibrosis Foundation Therapeutics Development Network.