We report results from an intervention study to improve communication during consultations about randomised clinical trials of cancer therapy. Patients, eligible for a trial, completed questionnaires about information preferences and attitudes to trials prior to seeing their doctors, who were either shown these questionnaires (intervention) or not (control). Fifteen doctors participated and invited 265 patients to join one of 40 different randomised clinical trials. Most patients (77.4%) agreed to trial entry and this was predicted by the Patient's Attitudes to Trials questionnaire with an 80.4% accuracy. Accrual, length of consultation, doctor and patient satisfaction were not associated with the intervention. Further research to explore the potential use of written interventions to facilitate communication and accrual to randomised clinical trials is recommended.