The aim of the study was to analyse the effect of posture on the relation between EMG amplitude and moment of the back muscles in different subjects, in order to gain a better insight into the possibilities of EMG as a means of measuring individual back load. Eight healthy subjects participated in the experiments. Isometric back extensions were performed in three postures: upright standing, standing with the spine flexed, and upright sitting. In each posture the lumbar moments of three maximal voluntary contractions were measured and then exertions of 10 to 90% of maximal voluntary contraction (increments of 10) were performed. EMG signals from the back muscles were recorded with four pairs of surface electrodes located 3 cm and 6 cm lateral to the midline of the spine at the L3 level. EMG signals were full-wave rectified and averaged over 2 s intervals of constant moment. The results show that posture has a considerable influence on the relationship between EMG and the lumbar moment. Besides this, large individual differences and an influence of electrode position on the relationship were found. Therefore it is concluded that whenever EMG is to be used as a means to measure individual back load, a calibration of the EMG amplitude to lumbar moment ratio is necessary for each subject, each electrode position, and each posture. Interpretation of EMG amplitudes has to be done on an individual basis and taking influences of posture and electrode location on the EMG-lumbar moment relationship into account.