Abstract Few studies have evaluated initiatives targeting implementation of cancer rehabilitation. In this study we aim to test the effects of a complex intervention designed to improve general practitioners' (GPs) involvement in cancer rehabilitation. Outcomes were proactive contacts to patients by their GP reported by the patients and GPs, respectively, and patients' participation in rehabilitation activities. Methods. Cluster randomised controlled trial. All general practices in Denmark were randomised to an intervention group or to a control group (usual procedures). Patients were subsequently allocated to the intervention or the control group based on randomisation status of their GP. Between May 2008 and February 2009, adult patients treated for incident cancer at Vejle Hospital, Denmark, were assessed for eligibility. A total of 323 general practices were included, allocating 486 patients to an intervention and 469 to a control group. The intervention included a patient interview about rehabilitation with a rehabilitation coordinator at the hospital, comprehensive information to the GP about individual needs for rehabilitation, and an encouragement to the GP to contact the patient proactively. Questionnaires were administered to patients and GPs at 14 months after inclusion. Results. At baseline average age of patients was 63 years and 72% were female. The most frequent cancer localisations were breast (43%), lung (15%), and malignant melanoma (8%). The intervention had no effect on either patient- or GP-reported extent of GP proactivity. Further, no effect was observed on patient participation in rehabilitation activities during the 14-month follow-up period. Discussion. The intervention had no effect on GP proactivity or on patient participation in rehabilitation activities. However, analyses showed a significant association between proactivity and participation and we, therefore, conclude that increased GP proactivity may facilitate patient participation in rehabilitation activities.