Abstract Objective This study was designed to describe the interventions made by pharmacists working within different therapeutic areas in medical practices in primary care and to estimate the effects on prescribing. Method All medical practices and community pharmacists in one health authority area in England were invited to participate. Pharmacists were allocated to work with specific practices and training was provided in musculoskeletal, respiratory and anxiolytic/hypnotic (sedative) prescribing. Pharmacists negotiated and agreed with their medical practice the type of interventions which would meet the practice's needs. Pharmacists were remunerated to provide a maximum of eight three-hour sessions in each medical practice on each therapeutic area over a 12-month period. Details of all interventions were recorded by each pharmacist and countersigned by a practice GP. A summary form was then sent to the health authority. Analysis of prescribing data before and after intervention in all practices was carried out using a predictive model to calculate estimated costs. Key findings Twenty-seven pharmacists working with 39 medical practices participated in the study. Pharmacist interventions resulted in 165 dose changes, 368 drug changes, 470 drug discontinuations and 1,042 patient reviews. On average, £3.99 was saved for each musculoskeletal intervention made. This compared with an additional cost of £15.50 and £17.92 for each intervention in sedative and respiratory prescribing, respectively. Analysis of prescribing data showed that the total spend in musculoskeletal and sedative prescribing in participating practices one year after intervention was lower than that predicted. Conclusion Accurate recording of individual interventions with reasons, outcomes and direct cost consequences allows purchasers to make more informed decisions about the potential benefits of practice pharmacists. This method may, however, underestimate the pharmacist's indirect impact on prescribing since it does not take into account any educational effect on prescribers.