This study sought to investigate the normal muscle fibre size and type distribution of the human erector spinae, both in thoracic and lumbar regions, in a group of 31 young healthy male (n = 17) and female (n = 14) volunteers. Two percutaneous muscle biopsy samples were obtained under local anaesthesia, from the belly of the left erector spinae, at the levels of the 10th thoracic and 3rd lumbar vertebrae. Samples were prepared for routine histochemistry for the identification of fibre types. Fibre size (cross-sectional area (CSA) and narrow diameter (ND)) was quantified using computerised image analysis. The mean CSA/ND for each fibre type was greater in the thoracic than the lumbar region, but there was no difference between the 2 regions either for percentage type I (i.e. percentage distribution by number), percentage type I area (i.e. relative area of the muscle occupied by type I fibres) or the ratio describing the size of the type I fibre relative to that of the type II. Men had larger fibres than women, for each fibre type and at both sampling sites. In the men, each fibre type was of a similar mean size, whereas in the women the type I fibres were considerably larger than both the type IIA and type IIB fibres, with no difference between the latter two. In both regions of the erector spinae there was no difference between men and women for the proportion (%) of a given fibre type, but the percentage type I fibre area was significantly higher in the women. The erector spinae display muscle fibre characteristics which are clearly very different from those of other skeletal muscles, and which, with their predominance of relatively large type I (slow twitch) fibres, befit their function as postural muscles. Differences between thoracic and lumbar fascicles of the muscle, and between the muscles of men and women, may reflect adaptive responses to differences in function. In assessing the degree of any pathological change in the muscle of patients with low back pain, it seems clear that (1) sex cannot be disregarded and (2) 'atrophied' (using the criteria from other muscles) type II fibres are not necessarily abnormal for the erector spinae, particularly in women.