Combined movement examination (CME) is used by physiotherapists to assess and treat patients with low back pain (LBP). However, this method has not been subjected to quantitative assessment. The purpose of this study was to discover if CME of the lumbar spine could be measured with acceptable intra-examiner reliability and to make a preliminary investigation of the effect of LBP on the results of CME. Combined movement examination of the human thoracolumbar spine was measured using a three-dimensional electromagnetic goniometer (3SPACE Fastrak Polhemus, Colchester, Vermont, USA). 1. Intra-examiner reliability of CME was evaluated using a test-retest design; examination was repeated in 23 subjects without LBP and 16 with LBP. 2. A sample of 31 subjects without LBP and 23 subjects with LBP underwent CME, which involved measuring right and left sideflexion in the flexed, neutral and extended position. Pearson's r and the intraclass correlation coefficient for all variables ranged from 0.79 to 0.93 (P < 0.05). A MANOVA test (P < 0.05), comparing the combined effects of the two variables for each position, was used to test for a difference between the positions attained by the subjects with LBP and those without. Generally LBP subjects showed smaller ranges of movement than those without LBP. Each position was recorded as a degree of rotation around the x-axis (flexion/extension) and around the y-axis (left and right sideflexion). CME can be carried out with acceptable intra-examiner reliability; preliminary evidence is presented concerning the effectiveness of CME in identifying reduced spinal movement in LBP subjects.