Background: Clinical ergonomic advice for sitting posture has been inferred from anthropometry theory and physiology studies. Qualitative observation of posture has been used to argue that postures are too flexed, too extended, too static, too mobile, require insufficient muscle activity or require too much. In other fields of healthcare, evidence is progressed from basic science to clinical trials before an intervention is prescribed, but postural advice has been prescribed from basic science without quantitative studies of postural behaviour. Spinal neuromuscular control can predict development of low back pain, it is affected by spinal pain, and can be trained, but studies examining these variables rarely measure the spinal position of their test subjects. There is a need for more detail of how spinal positions affect regional muscle activity. Study Objectives: The overall aims of this thesis were to quantify sagittal spinal postures in sitting for comparison within and between subjects and tasks, and to detail paraspinal and abdominal muscle activity associated with sitting postures. The studies quantified regional spinal curves in - Study I: typical posture behaviour during a computer task in comparison with standing, - Study II: postures that are achievable in sitting, - Study III: the regional muscle activity associated with sitting postures that have been clinically advocated as ‘ideal’, - Study IV: cohorts with and without a history of low back pain for comparison of regional muscle activity. Results and Conclusions: For clinical trials quantifying postural behaviour, the postural variation within subjects and task conditions in Study I demonstrated the importance of measurement over a prolonged period and subjects performing relevant tasks (as opposed to brief measures such as radiography that have been the standard for posture assessment). The use of surface tracking to quantify regional spinal curves and sagittal balance establish a foundation to investigate the effect of interventions on posture behaviour (eg. chair geometry, posture training, task variables and subject cohorts). Study I also showed that typical sitting posture for a computer task was more flexed at the thoracolumbar spinal region than when subjects deliberately ‘corrected’ their sitting posture, and both sitting postures were flexed at the lumbar region relative to standing. Study II showed that most subjects were unable to sit with spinal curves like those adopted in standing unless facilitation and feedback were provided, although these curves have been clinically advocated as ‘ideal’ posture. If clinical theories about ‘ideal’ sitting posture are correct, then teaching individuals the awareness of spinal position or skill to adopt these postures could be as important for workplace health and safety as other variables such as design and adjustment of office furniture. Study III showed three upright sitting postures that have been clinically advocated as ‘ideal’ were distinguished by incremental changes in activity of the lumbar multifidus muscles. In Study IV, individuals with a history of low back pain showed more incremental activity at the longissimus thoracis muscle to achieve the same sitting postures. If particular postures are shown to be ‘ideal’ in clinical trials, then training for these postures may need to focus on muscular strategies as well as spinal position. The distinct differences in regional muscle activity observed with spinal curves and subject cohorts (Studies III and IV) imply that studies of spinal neuromuscular control should measure or control spinal curves during testing. If spinal posture were controlled, the flat posture (flat surface from ~T5 to sacrum) would have the advantages of being achievable, commonly used in sitting and easily assessed. The flat posture also demonstrated the lowest muscle activity of the upright sitting postures examined, which may improve accuracy of determining muscle activity onset/offset used as an outcome measure for interventions, distinguishing cohorts or as a predictor for low back pain occurrence.