Objective To assess the effects of the abolition in April 2007 of prescription co-payments in Wales on rates of dispensing by general practices.

Design and setting An observational study comparing changes in dispensing in Wales, where co-payments were abolished, with North East England where they continued.

Participants All general practices in Wales and North East England which provided monthly dispensing data continuously between October 2003 and March 2008.

Main outcome measures The main outcome was the change in dispensing of all BNF items per 1000 patients per month between the baseline year before charges were first reduced in Wales and the year following abolition. Similar comparisons were made for the 14 medicines that had previously had most items dispensed subject to charge before abolition. Sales of over-the-counter substitutes were also examined where relevant. A survey of local health boards and primary care trusts examined local initiatives, which might differentially affect dispensing between the two areas.

Results Total monthly dispensing rates (items/1000 patients) increased significantly in both areas but significantly less so in Wales (difference=−19.7, p=0.024, 95% CI −36.7 to −2.6). For the 14 selected medicines, combined dispensing rates increased significantly in both areas but significantly more in Wales (difference=27.51, p <0.0001, 95% CI 23.66 to 31.35). There was much variation for individual drugs, but BNF categories tended to show this same trend except for antibiotics, where rates increased in Wales but decreased in NE England. The survey revealed few local initiatives that could explain these differences. Sales of over-the counter substitutes did not explain the changes in dispensing.

Conclusion The Welsh policy was associated with a smaller increase in dispensing of all BNF items than NE England. There was a small increase in dispensing rates relative to NE England for the 14 medicines with the highest number of items dispensed subject to charge before abolition. Although factors besides the co-payment may have influenced these changes, these results suggest that the overall impact of abolition was minimal.