In this study we examine what factors determine whether patients seen in a Veterans Administration primary care clinic will be referred to specialty clinics. It is based on a 25% sample of all patients seen by the medical residents in one clinic over an 18-month period. Our dependent variable is whether the patient was referred for a consultation or not. We performed separate regression analyses for patients who visited the clinic with acute complaints, with chronic-flareup complaints, and with chronic-routine complaints. Within each of these three categories, we further subdivided our sample into patients under and over 65. In each of the six regression equations we controlled for the medical diagnosis and for each resident's work load, residency year, and training. Work load and residency year had opposite (and unpredicted) effects for patients over the age of 65 compared to those under the age of 65. Higher work loads were associated with higher consultation rates for patients over 65 and with lower rates for those under 65. Residency year had a negative effect on referrals for patients over 65 and a positive effect for patients under 65.