Objective Recently the issue of access to health services has been brought into sharp focus by clear evidence of rationing patients queuing for NHS registration in the NHS General Dental Services. Conventional estimates suggest that about 50% of adults are registered per annum. This paper demonstrates that these conventional measures of access and utilisation can generate potentially misleading inferences. Design By analysing individual-level claims data from over 35,000 patients over 6 years we are able to: identify the underlying patterns of utilisation that generate the aggregate 50% registration rate; provide more detailed estimates of utilisation and access; and suggest possible determinants of the patterns of utilisation we observe. Setting Primary Care Results In contrast to conventional estimates of access we nd that access to GDS in Scotland is over 80% over 6 years. Moreover, we find that the population is comprised of a relatively large group of patients (30% of the population) who access GDP services at least once a year and a substantial group (19% of the adult population) who access services only once in 6 years. The groups who access services at intermediate frequencies are less numerous. Conclusions Assessing the e¤ectiveness of the public provision of health care services requires accurate information regarding access to those services. This paper sets out a framework for analysing and interpreting longitudinal data to provide information on the extent of access to health care services.