Standardized patients (SPs) are simulated patients or actual patients who have been carefully coached to present their illness in a standardized way. Much is known about the use of standardized patients in medical education. This article reviews advantages and disadvantages, reliability and validity of the use of standardized patients in general practice and primary care research. Performance in general practice can be measured with direct or indirect methods. With direct methods the physician-patient contact is directly observed or heard. Indirect methods are seldom complete and seldom accurate and therefore often invalid. Direct methods (observation, video, audiotapes, etc.) have face validity, but nevertheless have shortcomings. The SP method can mainly avoid the disadvantages of the other methods. The presentation of the case by the SP is accurate. The judgement of physician's behaviour during the consultation by the SP is accurate and reliable. SPs are generally believable. Less than one in five SPs is detected by the physicians, so the method has face validity. To obtain sufficient reliability and validity, a thorough selection and training of SPs is required, as is careful organization with an eye for detail. The SP method also has some important shortcomings. The method is time and work demanding, limiting the number of physicians that can be measured. In addition, measurement is usually limited to one consultation. In reality, however, diagnostic and therapeutic interventions are often spread over several consultations. This 'first-visit-bias' hampers conclusive answering of some research questions.