Objective: Complexity of care provided by general practitioners (GPs) is poorly addressed by current measures. Resource allocation for health care in inpatient settings utilises complex descriptions of clinical content and process to produce 'casemix' resource units. This pilot project explored the potential application of these concepts to the funding of general practice in Australia., Method: Prospective data were collected from three practices in the same community on one day in the same week. GP generated data included reasons for encounter, age and gender of the patient, whether the problems were new or old, what tests and referrals were initiated for each problem and global complexity ratings. Data collected from the patients included problems presented, problems not dealt with and satisfaction ratings. These data were used to explore the determinants of consultation complexity in general practice., Results: There were statistically significant correlations between: global complexity and problem type (coded to ICPC Chapter); global complexity and number of problems seen; and global complexity and consultation billing level. A resource allocation formula which considers broader descriptions than current fee descriptors is suggested., Conclusions: Aspects of the casemix concept may have some relevance to how resources could be allocated in general practice, but this would require substantial further investigation.