BACKGROUND: The NHS Choices website (www.nhs.uk) provides data on the opening hours of general practices in England. If the data are accurate, they could be used to examine the benefits of extended hours. AIM: To determine whether online data on the opening times of general practices in England are accurate regarding the number of hours in which GPs provide face-to-face consultations. DESIGN AND SETTING: Cross-sectional comparison of data from NHS Choices and telephone survey data reported by general practice staff, for a nationally representative sample of 320 general practices (December 2013 to September 2014). METHOD: GP face-to-face consultation times were collected by telephone for each sampled practice for each day of the week. NHS Choices data on surgery times were available online. Analysis was based on differences in the number of surgery hours (accounting for breaks) and the times of the first and last consultations of the day only between the two data sources. RESULTS: The NHS Choices data recorded 8.8 more hours per week than the survey data on average (40.1 versus 31.2; 95% confidence interval [CI] = 7.4 to 10.3). This was largely accounted for by differences in the recording of breaks between sessions. The data were more similar when only the first and last consultation times were considered (mean difference = 1.6 hours; 95% CI = 0.9 to 2.3). CONCLUSION: NHS Choices data do not accurately measure the number of hours in which GPs provide face-to-face consultations. They better record the hours between the first and last consultations of the day.