SummaryEffect of the introduction of the problem-oriented record (POR) system into a military clinic was compared with the effect of the traditional source-oriented record (SOR) on the following variables measured for appointment patients: waiting time, time spent with physician, amount of data collected, number of problems identified, data collected per min, and number of problems per min.The POR system consisted of a standardized data base collected with the assistance of a specially trained secretary, a problem list, problem-oriented progress notes, and a biweekly peer record audit. Of 241 appointment patients seen, 60 control and 75 POR records were retrieved for grading. Records were graded according to preset criteria for amount of data collected. The typical reliability of a single grade, calculated as the intraclass correlation, was 0.78. Test retest reliability (correlation coefficient) for the principal grader was 0.95.Multivariate analysis of variance of the simultaneous effects of the record format (POR or SOR), patient status (new or followup), and physician on the dependent variables-amount of data collected, number of problems identified, data collected per min and number of problems identified per min-indicated that the POR system increased significantly these dependent variables. However, the POR had no effect on mean waiting time (POR 31.7 & , SOR 30.2 min) or on mean time spent with the physician (POR 15.9 min; SOR 17.1 min), but did increase time spent with followup patients and decreased time spent with new patients. New patients had significantly more data collected, more problems identified, and more time spent with them than did followup patients. Cbange in physician only affected the amount of time spent with the patient.It was concluded that in a military clinic the POR system can increase the number of problems identified, and the amount of data collected, without increasing physician time spent with the patient. Between 1916 and 1920, Codman (4) first demonstrated that surgical care improves if surgeons keep careful record abstracts that include immediate and long-term outcomes. Codman's idea fell into disuse until the mid-1950's, when Lembke (9) and others (14, 15), began to use consistent criteria to measure quality of care as reflected in the medical record. These studies reaffirmed the fact that the existence of a record that can be audited can improve medical care. In 1964, Weed (20) proposed that changing from a traditional source-oriented record (SOR) to a problem-oriented record (POR) system (i.e., the POR format + regular audit) would improve the process of patient care. Although the POR has been adopted widely (I), Weed's proposal has elicited much debate (5,8,19). However, only a handful of studies have attempted to demonstrate whether or not introduction of the POR affects patient care (2,3,6,7,18). The results of these studies are inconclusive. Whereas a few researchers suggest that the POR increases both the amount of data collected and its accessibility, and decreases...