Background Cancer often affects the quality of life and well-being of patients as well as their relatives. Previous studies have suggested that relatives should be involved in psychosocial rehabilitation to address the needs for an interpersonal relationship with others in the disease trajectory. We developed an innovative rehabilitation program to be offered to the patient and a relative as a pair. Objective The aim of the present pilot study was to examine the feasibility of the intervention in a randomized controlled trial (RCT) and to evaluate the impact on quality of life. Methods The study was designed as an RCT comparing the new multimodal psychosocial rehabilitation with the usual services. The intervention comprised three ‘supportive talks’ and a residential rehabilitation course. From March 2010 to March 2011, participation was offered at the time of diagnosis to patients with lung or gynecological cancer from two departments at Odense University Hospital in Denmark. Questionnaires were used to estimate changes in quality of life (EORTC-QLQ-C30 on global health status) and well-being (WHO-Five Well-Being Index) at baseline and after 2 and 12 months. Information on the participants’ views about the rehabilitation intervention was obtained from assessment charts and qualitative interviews. Results A total of 209 patients were assessed for eligibility, but only 42 pairs were randomized to the study. The 2-month follow-up was completed by 34 patients and 32 relatives, and 19 patients and 21 relatives completed the 12-month follow-up. A higher dropout rate at the 12-month follow-up was reported in the intervention group compared with controls. Quality of life and well-being increased for patients and relatives in both the intervention and the control group, and no clinically significant difference was observed between the intervention and the control group. Pairs reported that the time of inclusion was inconvenient and that rehabilitation ought to meet their changing needs. Conclusions The pilot study showed that it may be difficult to conduct an RCT of a psychosocial rehabilitation intervention for pairs, and difficulties with inclusion and drop out have to be addressed. Interventions need to be carefully developed and tested before evaluating an effect in a large-scale study.