In all the models, with or without morbidity measures, patients in higher deprivation deciles have greatercost (Table 7).Theuncon- ditional cost ratio between the 10th and 1st deprivation deciles is 1.50 and all the conditional cost ratios between the 10th and 1st deprivation decile from modelswhich include a morbidity measure are also statistically significantly greater than 1.00. Our results sug- gest that there is horizontal pro-poor inequity in primary care even after allowing for rich clinical morbidity measures.