A regular clinical pharmacy service to wards is well established in Britain, yet there has been little practice research in this area. In this study, we established how pharmacists distribute their time between different activities on all the wards. All 28 pharmacists in two London hospitals were observed. The observer recorded their activities at one minute intervals and categorised the activity. Thirty‐three hours and five minutes of observations were made. The pharmacists visited a total of 55 wards, and checked 982 drug charts.The major activities were prescription monitoring, which accounted for 31 per cent ± 2 per cent, of the time (mean of pooled data ± 95 per cent confidence intervals), travel to wards 21 per cent ± 2 per cent, stock control 12 per cent ± 1 per cent, transcription 8 per cent ± 1 per cent and clinical monitoring (such as checking giving sets) 7 per cent ± 1 per cent.The proportion of time spent on each activity was very similar between the two sites, except for the category “transcription”, which took 11 per cent ± 2 per cent of the time in one hospital, where all discharge prescriptions were transcribed onto ward pharmacy sheets (allowing processing in the pharmacy without the need for drug charts to leave the ward), and 3 per cent ± 1 per cent at the other site, in which they were not transcribed.These results demonstrate that, when on the ward, pharmacists spend the majority of their time in a clinical role. The methodology is simple to use and appears reproducible. This work, if repeated locally, could be used to highlight areas of inefficiency, and could be used to test and predict the impact of other methods of delivery of clin