Variability in maximal expiratory flow volume (MEFV) curves was studied in a group of non-smokers and asymptomatic smokers, using the body Plethysmograph to measure lung volume. Several factors were evaluated for their ability to predict maximal expiratory flows (MEF) at various lung volumes. Of the factors studied, age, sex, and peak expiratory flow (PEF) were shown to be of value. Using age and sex, the mean coefficient of variation for MEF was 27 per cent for several lung volumes expressed as per cent of vital capacity (VC). Adding PEF as a predictor decreased the variability. Relating flow to lung volume expressed as per cent of total lung capacity (TLC) and expressing flow as VC per second or TLC per second were not helpful. Further insight into the factors influencing the variability in MEFV curves is necessary before the ultimate value of these curves in detecting mild obstructive lung disease can be realized.