OBJECTIVES: To determine how women and physicians rate individual characteristics of contraceptives. METHODS: Discrete choice experiments are used in health economics to elicit preferences for healthcare products. A choice experiment uses hypothetical scenarios to determine which individual factors influence choice. Women and general practitioners (GPs) were shown individual characteristics of contraceptives, not always matching existing methods, and chose the best and worst features. RESULTS: Two hundred women, mean age 36, 71% using contraception, were presented with descriptions of 16 possible methods and asked to indicate their preference for individual characteristics. One hundred and sixty-two GPs, mostly women, also completed 16 descriptions. Longer duration of action was most favoured by both, followed by lighter periods with less pain or amenorrhoea. The least attractive features for women were heavier and more painful periods, high cost, irregular periods, low efficacy (10% failure) and weight gain of 3 kg. GPs ranked a 10% pregnancy rate as least attractive followed by heavy painful periods and a 5% failure rate. CONCLUSION: Women and GPs differed in their ranking of contraceptive characteristics. Long duration of use, high efficacy, minimal or no bleeding without pain, were preferred by both. Very undesirable were heavy periods especially with pain, and low efficacy.