Objective criteria for interpretation of peak expiratory flow rate readings were assessed in 50 patients evaluated for suspected occupational asthma who had at least two weeks of PEFR readings and an objective diagnosis based on other investigations. The prevalence of OA was 36 percent. Peak flows were interpreted by two observers blinded to other results. Criteria for a PEFR interpretation of OA were as follow: diurnal variation greater than or equal to 20 percent relatively more frequently or with greater variation on working days than days off work. With the objective diagnoses as the gold standard, the sensitivity of the PEFR interpretations was 72 percent for OA; specificity for no asthma was 53 percent. Excluding those with greater than or equal to 20 percent variation on only one day sensitivity improved to 93 percent for OA, and specificity to 77 percent. There was an acceptable level of interobserver variation (kappa 62 to 83 percent). We conclude that simple objective criteria for PEFR interpretation can be developed with acceptable interobserver variation.