A protocol for upper-respiratory tract complaints was administered to 226 patients in a walk-in clinic. The protocol, for use by a physician-extender in conjunction with a physician, specified the collection of data necessary for management. A decision-making algorithm separated the major causes of upper respiratory infection (URI) complaints and led to one of four plans: a physician referral, a culture only, antibiotic treatment, or symptomatic treatment only. Each patient was seen by a physician following the health assistant's interview. Of 226 patients, 96 (42%) would have been sent home by the protocol without seeing the physician. None of these had a complication of URI. Sixteen (7%) of the 226 had serious complication— all would have been referred to the physician. The protocol proved to be safe and efficient, acceptable to patients, and a reliable approach to physician-extender management of URI.