A Patient Panels Registry was developed to study patient identification with family physicians in a large health maintenance organization. There was a clear correlation between the sex of the patient and the sex of the physician. Female patients were 1.49 times as likely as males to select a female physician. Male patients were 1.14 times as likely as females to select a male physician. Women physicians were found to have panels consisting of 66.4 percent female patients, while panels of the male physicians were 53.8 percent female. In order to interact most effectively with both male and female patients, family physicians must be able to recognize their own sex biases as well as those of their patients.