Three thermometers, mercury in glass, Becton-Dickinson digital and IVAC tympanic membrane thermometer, were compared. The study was designed to test the null hypothesis that there is no difference between the thermometer recordings, meaning that electronic thermometry is a suitable alternative to traditional mercury in glass thermometry. Children aged between 0 and 15 years old were entered into the study, n = 114. Standard procedures were used throughout the study to minimize the risk of errors in the data collection and technical errors were recorded. The data were analysed using graphical techniques described by Bland and Altman. The differences between the temperature recordings were plotted against their mean (the estimated true value), with the bias and limits of agreement (2 standard deviations from the mean) calculated for the mercury in glass/digital, mercury in glass/tympanic and digital/tympanic temperature recordings. The range of temperatures were 35.1 degrees C-38.5 degrees C, with a mean of 36.85 degrees C for the mercury in glass thermometer, 36.8 degrees C for the digital thermometer and 36.65 degrees C for the tympanic thermometer. The bias for each comparison was 0.36 degrees C, 0.21 degrees C and 0.17 degrees C and the limits of agreement were wide, -0.516 degrees C to 1.234 degrees C, -0.84 degrees C to 1.252 degrees C and -0.940C to 1.244 degrees C for the mercury in glass/digital, mercury in glass/tympanic and digital/ tympanic recordings respectively. A difference of 0.2 degrees C is usually accepted for clinical practice, therefore the null hypothesis was rejected.