The aim of this thesis is to determine whether surface electromyography (EMG) can be used as a diagnostic tool in chiropractic practice to identify the functional status of the lumbar paraspinal muscles. There were two main studies to achieve this aim. The reliability and validity of the surface EMG signal to measure the activity of paraspinal muscles during maintenance of simple static postures was evaluated. During maintenance of static postures, the raw surface EMG signal was often contaminated by an electrocardiographic (ECG) signal. Although the ECG artefact was successfully removed using two different ECG removal techniques (manual and semi-automatic), the reliability of the surface EMG signal was not significantly improved (ICC less than 0.75) for both non-normalised and normalised data. Therefore the static postures that were used in this thesis did not provide a protocol that can be used to measure the functional status of the lumbar paraspinal muscles in clinical practice. However, when muscle contraction was at a moderate level, the reliability of EMG signal became better. Walking was considered to be a possible protocol to record a reliable surface EMG signal from paraspinal muscles. Three components of the surface EMG signal were used to characterise the pattern of muscle activity during steady state walking. The narrow window technique was used to characterise the peak activation point of the activity envelope in order to capture a stationary signal from which to calculate amplitude and frequency measures. Walking is a cyclic activity. The back muscles contract rhythmically during a single gait cycle. It is possible to identify the start and end points of the activity envelope associated with the rhythmic contraction of the muscles and define the timing of the muscle activation cycle relative to heel strike. The metronome was found to be useful to control the pace of natural walking in this study. The surface EMG signal of the first recording minute (1 ~ 2 minute) was not associated with a signal that was stable in terms of the parameters that were used in this study. It wa s found that the last recording minute (9 ~ 10 minute) can be used. This suggests that it may be necessary for subjects to walk for a defined period lasting some minutes before the commencement of recording of the surface EMG. Surface EMG may be used as a tool to measure activation patterns of the low back muscles during muscle contraction associated with the support of various static postures or during the execution of dynamic movements such as walking in the real world. The static postures used in this thesis to record the surface EMG signal from the lumbar paraspinal muscles were found not to provide the basis for a reliable and valid tool. However, a walking exercise might be an alternative activity which can be used easily in clinical practice. The components of the surface EMG signal that may be used in future studies might include measures of the amplitude, frequency and timing of the surface EMG signal.