Objectives: to examine current practice with regard to hydration in acute stroke (part 1) to identify problem areas, to develop, implement and evaluate an intervention package (part 2). Design: prospective, repeated measures, between subjects Setting: a district general hospital. Participants: consecutive admissions with an acute stroke. Part 1 involved 30 patients, 14 of whom were women whose median age was 73 (45–90), part 2 involved 48 patients, 27 of whom were women whose median age was 73 (47–91). Main outcome measures: amount of fluid prescribed, received and mode of delivery in the first 2 weeks plus standardized swallowing assessments on admission, days 3, 7 and 14. Results: in both parts 1 and 2 of the study patients were appropriately managed according to the results of the SSA. However, in part 1 it was found that patients were documented as, on average, receiving less than 21 (regardless of mode of delivery) and were being prescribed little more than 1.51 per day. In part 2 there was a 33% increase (P<0.005) in fluid prescribed, but no significant increase (P=0.76) in the amount of fluid documented as having been delivered. This was independent of admitting ward. Conclusions: the intervention package had little or no effect on practice. The reasons are discussed and recommendations made. Further interventions are to be explored, implemented and reaudited in the future