PURPOSE: Machine learning-based methods, which include Artificial Neural Networks [ANN], have been used successfully in varied classification problems. If these methods can successfully classify those vulnerable to musculoskeletal problems such as low-back pain [LBP], they may have utility in screening and management of such conditions and aid in identifying what assessment methods provide optimal information for practitioners. We examined whether ANN techniques could correctly classify whether subjects experienced LBP in a convenience sample of dancers. METHODS: 60 subjects [48 women], 36 of whom [24 women] reported an episode of back pain in the past two months, were instrumented with a multi-segment spine marker set [Swain et al., 2019] and recorded [Motion Analysis Corp Eagle, 250 Hz] while performing standing and seated rotations, walking and several functional and dance-related movements [step-over task, arabesque, passe]. The multi-segment model has five segments: pelvis, lower and upper lumber [L/UL], lower and upper thoracic [L/UT]. 3D rotations were computed both between adjacent segments, and with respect to the lab coordinate system. To determine variables of interest, 1-dimensional statistical parametric mapping [SPM1D; Pataky, 2008] analysis was performed. Features [min, max, time to min/max, and side-side difference] were extracted from these variables and used to train an ANN pattern recognition tool [MATLAB]. Approximately 75% of the data were used for training, with the remainder used for validation and testing. Because of the dearth of men, analysis was performed on the entire cohort, and of women only. RESULTS: Based on the SPM1D analysis, only approximately 10% of data were used for training the ANN. For example, for walking trials, LL and LT axial and UT coronal plane rotations were used. The ANN classifier was able to correctly identify incidence of LBP with approximately 65% accuracy. CONCLUSIONS: Based on our small sample, ANN techniques show promise for identifying subjects with LBP based on their movement patterns. A larger training set of data is needed for better results. Future work should optimize feature selection by focusing on areas of difference between data rather than by selecting fixed features [e.g., max value] and examine the effect of different ANN architectures.