Abstract Although lumbar support belts (LSBs) are not effective at preventing a first episode of low back pain (LBP), the advancement of knowledge now favours their use by workers who already have this kind of pain. On the assumption that a delayed return to work risks harming workers’ health, the option of wearing an LSB may potentially be beneficial during a gradual return to work, since it could provide psychological support to reassure the patient. Thus, this could represent an additional tool among the various methods used to intervene with the psychosocial factors related to individuals and their work environments in order to facilitate a return to work and also help workers stay at work. Nevertheless, it is also important for the LSB to provide biomechanical support in the sense of mechanical stability for the lumbar spine in order to enhance workers’ safety. Two kinds of flexible belts are sufficiently comfortable to be used in the workplace: (1) elastic belts (E-LSBs); and (2) non-elastic belts (NE-LSBs). The purpose of this study was to compare these two classes of LSBs in terms of the psychological and biomechanical support they provide for both healthy subjects and subjects with LBP. On an exploratory basis, it was also relevant to verify whether different subgroups of subjects with LBP, classified on the basis of biomechanical (lumbar instability) or psychological (fear of movement or pain) hypotheses, would experience different biomechanical effects. With a view to assessing the effect of wearing an LSB in a potential clinical trial, it was necessary to evaluate certain variables associated with the adoption of this type of intervention, also on an exploratory basis. Two groups of subjects, divided equally according to sex, were compared: 20 healthy subjects and 40 subjects with LBP; the latter showed substantial variations with regard to signs of lumbar instability (biomechanical dimension) and fear of movement (psychological dimension). Two sessions in the laboratory enabled us to assess different psychological and biomechanical parameters in three experimental conditions (no LSB, E-LSB, NE-LSB). Session 1: (1) pain (subjects with LBP); (2) lumbar proprioception, measured with an instrumented chair that allows for movement (10°) in axial rotation (torsion) of the lumbar region; (3) postural control of the trunk on an unstable chair; (4) lumbar spine rigidity, measured with a device that generates mild jolts (front and back) to the trunk. Session 2: (with measurements of muscle activation and movement): (5) anticipatory postural adjustments (preactivation of trunk muscles preceding an anticipated disturbance) to a rapid flexion of the right arm; (6) lumbar muscle coordination and range of motion during three standardized tasks. These three tasks were lifting and putting down crates, maximum trunk flexions forward and back to the vertical position, and slight (±20°) trunk flexions and extensions close to the neutral (vertical) position. During both sessions, LBP subjects were asked questions about their fears related to pain after executing some of these tasks, namely the ones that pose the greatest threat to the lumbar spine. On an exploratory basis, LBP subjects were also questioned, at the end of session 2, about some factors presumed to be associated with wearing an LSB, namely favourable attitudes to wearing an LSB, their feelings about its functional effectiveness, and anticipated stigmatization by the people around them. The data gathered in the laboratory were analyzed not only to compare the three experimental conditions (no LSB, E-LSB, NE-LSB) but also to compare the groups (healthy subjects and LBP subjects) and, even more importantly, to compare subgroups of LBP subjects identified on the basis of measures that are assumed to be associated with lumbar instability or fear of movement/pain, given the biomechanical and psychological hypotheses that promote the wearing of an LSB. Several immediate effects of LSB use were studied, but they were equivalent for both types of flexible LSBs (E-LSB and NE-LSB). In fact, neither the measures associated with pain nor the biomechanical measures nor the measures potentially associated with deciding to wear an LSB nor even the preference for one LSB or the other determined which belt would have positive effects. Thus, both LSBs appear to provide the same benefits, at least with regard to immediate effects. On the mechanical level, these results also suggest that the NE-LSB does not produce more intra-abdominal pressure than the E-LSB, due to its lack of elasticity during disturbances to or changes of trunk posture. Both LSBs resulted in decreased pain in a standing position and decreased fear and dramatization of pain during various tasks considered to endanger the lumbar region, especially those that apply a greater load to spinal structures. Regarding the biomechanical variables, many similar effects were observed and demonstrated in the healthy subjects and the LBP subjects. First of all, the impacts on the variables related to motor control mechanisms were mixed: that is to say, there was no effect on lumbar proprioception and low-level effects that were sometimes favourable and sometimes unfavourable on anticipatory postural adjustments and postural equilibrium. The biomechanical variables associated with mechanical effects produced clearer positive impacts with regard to an increase in lumbar spine rigidity, a decrease in maximum lumbar flexion and a reduction in trunk muscle activation. Finally, functionally speaking, the effects during the task that involved lifting and putting down crates were also positive, as they were associated with a reduction in lumbar flexion without, however, a negative impact on the movement of adjacent segments such as flexion of the thorax, pelvis or knees. Overall, exploration of the possible existence of clinical subgroups, formed on the basis of biomechanical (lumbar stability) and psychological (fear of pain or movement) hypotheses was not conclusive but remains a possibility to consider in a potential clinical trial studying the long-term effects. On the other hand, the exploratory study of the variables that are presumed to be associated with wearing an LSB, namely favourable attitudes to wearing an LSB, a feeling of functional effectiveness and anticipated stigmatization, as measured in the LBP subjects, produced encouraging results, regardless of participants’ age and sex. These results even lead us to believe that favourable attitudes to the wearing of an LSB may be associated with reduced fear and dramatization of pain during tasks considered to be harmful for the back. To sum up, the psychological variables associated with pain and those that are presumed to be associated with wearing an LSB support the hypothesis that wearing such a belt can reassure patients with LBP, which in turn could favour continuation at work during periods of recurrence or speed up the return to work after periods of absence. For their part, the biomechanical variables indicate that the mechanical effects have the potential to make this practice safe. A randomized clinical trial targeting secondary (prevention of medical leave) and tertiary (prevention of long-term leave) prevention is necessary to verify these hypotheses with a population of workers in the subacute phase of their LBP or workers with recurring pain.