Objective: we wish to determine the constructs used by patients to evaluate conventional and complementary medicine (CAM), prioritise these constructs and determine how these characterise patients‘ choice for CAM and conventional medicine. Methods: a repertory grid technique was employed, in interviews of 20 patients, 10 from a general practice and 10 from a CAM clinic. Results: CAM was used with greater frequency in chronic illness and generated more constructs than conventional medicine, although the constructs were similar. The priorities for these constructs were different: specificity of treatment and severity of illness were prioritised for conventional medicine and lack of science was prioritised for CAM. The CAM characteristic most commonly cited was non-chemical. Conclusion: this exploratory initial study suggests that people use very similar criteria to evaluate CAM and conventional medicine. Their choice of treatment was based largely on the type of illness from which they were suffering. Further research is needed in this area.