Different back muscle reflex assessment protocols have shown abnormally longer reflex latency responses of back muscles in chronic low back pain (CLBP). However, many confounding variables are difficult to control, such as the load magnitude and the preactivation of trunk muscles. The aims of this study were to evaluate, in 30 subjects with CLBP and 30 healthy controls, the activation levels of back muscles during pre-loading and their reflex responses to sudden loading. After subjected to six practice perturbations, 20 sudden and unexpected forward perturbations of the trunk were applied in 30 CLBP subjects (14 women) and 31 controls (17 women), while attempting to minimize the confounding effect of preactivation level and perturbation amplitude. Reflex latency and amplitudes were computed from the surface EMG signals of four back muscles (bilaterally at L5, L3, L1, T10 vertebral levels). EMG was also collected from abdominal muscles. Subjects with CLBP significantly increased the preactivation of back muscles (abdominal preactivation the same) relative to controls while no sex effect was observed. While adjusting statistically for these differences, reflex amplitude was significantly higher in subjects with CLBP and men, compared to healthy controls and women, respectively. Interestingly, contrary to most of the literature available, no between-group effects were detected for reflex latency, which could potentially be explained by an appropriate control of confounding variables, but this remains to be clarified in future research.