Objective To compare the effectiveness of pain neuroscience education (PNE) versus biomedical education (BME) in a rehabilitation program following arthroscopic rotator cuff repair (ARCR) in patients with chronic shoulder pain. Methods Twenty-nine patients who participated in a rehabilitation program were randomly assigned to either an experimental PNE group (N = 16) or a control BME group (N = 13). Measurements included pain intensity at rest and in movement, pain catastrophizing, kinesiophobia, disability and health-related quality of life (HRQoL). Outcomes were evaluated at baseline and at 4 and 8 weeks after the intervention. Results A main effect for time was observed for: intensity of pain at rest (p < .01); pain with movement (p < .01); pain catastrophizing (p < .01); kinesiophobia (p < .01); disability (p < .01); and HRQoL (p < .01). No group interactions were significant for any variable, except for pain with movement, which favored the PNE group (p = .03). Large effect sizes (ranging from d = 0.79 to d = 2.65) were found for both interventions in all outcomes. Conclusion A rehabilitation program including either PNE or BME are equally effective in improving rest pain, pain catastrophizing, kinesiophobia, disability, and HRQoL in patients after ARCR, except for pain at movement in favor of the PNE group. The inclusion of PNE in the rehabilitation program appears to lead to clinically meaningful improvements in pain at rest in short term when treating patients with ARCR.