Personal heat strain monitoring provides the best means for maximizing worker safety and productivity in hot jobs. The present study compared methods for reducing the environmental effect on aural canal temperature in an attempt to reduce the predictive error of a personal heat stress monitor (QST) and a simple ear thermistor (EAR). Subjects underwent three exposures in impermeable protective clothing (PC) in an environment of 30.1 degrees C wet bulb globe temperature (WBGT) wearing either ear plugs (PLG), ear moulds (MLD) or ear moulds and earmuffs (MFS). Mean work time across all trials was 63.1 +/- 7.9 min, with a mean rectal temperature at 60 min of 38.5 +/- 0.1 degrees C and a sweat production rate of 29.7 +/- 8.6 g/min. Rectal temperature was used as the criterion measure of core temperature. Although group mean predictions were satisfactory, large SD of mean differences (+/- 0.36) meant that predictive ability of QST for individuals was greatly impaired. A simple well-insulated ear thermistor showed slightly better accuracy (+/- 0.28) as a predictor of rectal temperature across time, particularly at peak temperatures. Incorporating the predictive error of aural temperature still extended the duration of safe work times relative to ACGIH guidelines. Further investigation is still necessary to ensure that predictions based on aural temperature are accurate and safe for the worker population across varying conditions.