Background: Shoulder symptoms are common in patients with hypermobility spectrum disorders (HSD), but few studies focus on identifying factors associated with treatment effects. Aim: To identify baseline and clinical characteristics associated with a better outcome 16 weeks after starting an exercise-based treatment in patients with HSD and shoulder symptoms. Design: Exploratory secondary analysis of data from a randomised controlled trial. Method: Self-reported treatment outcome was reported as change between baseline and follow-up after 16 weeks of high-load or low-load shoulder strengthening. Multiple linear and logistic regressions were used to investigate associations of patient expectations of treatment effect, self-efficacy, fear of movement, and symptom duration with change in shoulder function, shoulder pain, quality of life, and patient reported health change. All regression models were performed firstly with adjustments for covariates (age, sex, body mass index, hand dominance, treatment group, and baseline score of the outcome variable) and secondly with additional adjustments for exposure variables. Results: Expectations of complete recovery were associated with an increased odds of perceiving an important improvement in physical symptoms after a 16-week exercise-based treatment program. Higher self-efficacy at baseline seemed to be associated with improved shoulder function, shoulder pain and quality of life. A higher fear of movement seemed to be associated with increased shoulder pain and decreased quality of life. A longer symptom duration was associated with decreased quality of life. Conclusion: Expectations of complete recovery, higher self-efficacy, lower fear of movement and shorter symptom duration seem to be important for better treatment outcomes.