Peak expiratory flow (PEF) monitoring is often used to establish the relationship between occupational exposure and asthma. FEV1 has been found to be a better physiologic index than PEF in the measurement of airflow obstruction. The aim of this study was to compare the accuracy of serial monitoring of PEF and FEV1 in the diagnosis of occupational asthma. Twenty consecutive subjects referred for possible occupational asthma were asked to perform serial monitoring of PEF and FEV1 using a portable ventilometer. Two sets of graphs were plotted for both PEF and FEV1: graphs with the best of all values and graphs with the best of two reproducible values. Three observers interpreted both PEF and FEV1 recordings by the visual method in a blind, randomized manner as either compatible with occupational asthma or not. Eleven of the subjects had a positive inhalation challenge test (high-molecular-weight agents, n = 6; low-molecular-weight agents, n = 5). In the case of analysis of the graphs plotted with the best of all values, the sensitivity of the PEF recording interpreted by the three observers was 82, 73, and 73%, and of the FEV1 recording as 55, 55, and 45%; specificity of PEF recording was 89, 100, and 100%, and of FEV1 was 56, 89, and 100%. When an agreement between two of the three readers was required to define occupational asthma, sensitivity and specificity were 73 and 100% for PEF and 55 and 89% for FEV1. Lower sensitivities were found when the same analyses were performed with the graphs plotted with the best of two reproducible values. It was concluded that unsupervised FEV1 is not more accurate than unsupervised PEF monitoring in the diagnosis of occupational asthma. Plotting graphs using the best value gives better diagnostic accuracy than plotting them with the best of two reproducible values.