Background: The combined use of genetic markers and detectable minimal residual disease identifies patients with chronic lymphocytic leukaemia with poor outcome after first-line chemoimmunotherapy. We aimed to assess lenalidomide maintenance therapy in these high-risk patients. Methods: In this randomised, double-blind, phase 3 study (CLLM1; CLL Maintenance 1 of the German CLL Study Group), patients older than 18 years and diagnosed with immunophenotypically confirmed chronic lymphocytic leukaemia with active disease, who responded to chemoimmunotherapy 2-5 months after completion of first-line therapy and who were assessed as having a high risk for an early progression with at least a partial response after four or more cycles of first-line chemoimmunotherapy, were eligible if they had high minimal residual disease levels or intermediate levels combined with an unmutated IGHV gene status or TP53 alterations. Patients were randomly assigned (2:1) to receive either lenalidomide (5 mg) or placebo. Randomisation was done with a fixed block size of three, and was stratified according to the minimal residual disease level achieved after first-line therapy. Maintenance was started with 5 mg daily, and was escalated to the target dose of 15 mg. If tolerated, medication was administered until disease progression. The primary endpoint was progression-free survival according to an independent review. The pre-planned interim analysis done by intention to treat was done after 20% of the calculated progression-free survival events. This study is registered with ClinicalTrials.gov, number NCT01556776; treatment in the lenalidomide group is still ongoing. Findings: Between July 5, 2012, and March 15, 2016, 468 previously untreated patients with chronic lymphocytic leukaemia were screened for the study; 379 (81%) were not eligible. Recruitment was closed prematurely due to poor accrual after 89 of 200 planned patients were randomly assigned: 60 (67%) enrolled patients were assigned to the lenalidomide group and 29 (33%) to the placebo group, of whom 56 (63%) received lenalidomide and 29 (33%) placebo, with a median of 11·0 (IQR 4·5-20·5) treatment cycles at data cutoff. After a median observation time of 17·9 months (IQR 9·1-28·1), the hazard ratio for progression-free survival assessed by an independent review was 0·168 (95% CI 0·074-0·379). Median progression-free survival was 13·3 months (95% CI 9·9-19·7) in the placebo group and not reached (95% CI 32·3-not evaluable) in the lenalidomide group. The most frequent adverse events were skin disorders (35 patients [63%] in the lenalidomide group vs eight patients [28%] in the placebo group), gastrointestinal disorders (34 [61%] vs eight [28%] ), infections (30 [54%] vs 19 [66%] ), haematological toxicity (28 [50%] vs five [17%] ), and general disorders (28 [50%] vs nine [31%] ). One fatal adverse event was reported in each of the treatment groups (one [2%] patient with fatal acute lymphocytic leukaemia in the lenalidomide group and one patient (3%) with fatal multifocal leukoencephalopathy in the placebo group). Interpretation: Lenalidomide is an efficacious maintenance therapy reducing the relative risk of progression in first-line patients with chronic lymphocytic leukaemia who do not achieve minimal residual disease negative disease state following chemoimmunotherapy approaches. The toxicity seems to be acceptable considering the poor prognosis of the eligible patients. The trial independently confirms the clinical significance of a novel, minimal residual disease-based algorithm to predict short progression-free survival, which might be incorporated in future clinical trials to identify candidates for additional maintenance treatment. Funding: Celgene Corporation. © 2017 Elsevier Ltd.