This paper reports an evaluation of a counselling service that was introduced into 20 general practices within Bradford City Primary Care Trust. Clinical Outcomes in Routine Evaluation (CORE) Assessment and End of Therapy forms were used to record referral and attendance information along with problems identified and demographic information about clients. There was a large disparity in the number of referrals across practices illustrating a range of influences on take-up of the service. Levels of attrition from referral for counselling to commencing were relatively high (188 invitations: 89 took up counselling). A longitudinal evaluation of outcomes was conducted using the CORE outcome measure and the SF-36 Health Survey Questionnaire. Data was collected at the beginning of counselling, at six months and at 12 months follow-up. We found some indication of positive change relating to severity of problems and ability to engage in social activities. However, evaluators should be aware that high attrition rates can create problems in achieving statistical reliability in contexts where overall population numbers are relatively small and/or the time for data collection is relatively short. In light of the recent growth of counselling within primary care, and the need to evaluate these services, the issues encountered in setting up this counselling service, and in seeking to evaluate it, may be instructive to others. Focus groups with counsellors and practice staff explored the benefits and problems in providing the service within practices.