BackgroundConsumption of opioids is increasing worldwide in people with chronic non-cancer pain, although their effectiveness is debated. ObjectivesThe aim of the current study was to evaluate analgesic consumption and its association with different variables (demographic variables, pain, anxiety/depression, catastrophism, and kinesiophobia), in the field of musculoskeletal rehabilitation, where no data are available. MethodsThis was a retrospective study over a period of 8 years on people hospitalised for rehabilitation after injury. Participants were classified into 3 categories: no analgesics (NA), non-opioid analgesics (NOA), and opioid analgesics (OPA). ANOVA or chi-squared tests were used to compare the 3 groups. ResultsA total of 4,350 people (84% men; mean [SD] age, 44 [11] years) were included. In total, 20% were taking OPA, 40% NOA and 40% NA. In the OPA group, tramadol was mainly used, and the morphine equivalent median dose was 8.3 mg/day. In the NOA group, paracetamol and ibuprofen were mostly used. Symptoms increased progressively across the 3 groups (NA/NOA/OPA), with increased levels of pain severity/interference, anxiety/depression and catastrophizing, and a higher prevalence of neuropathic pain in the OPA group versus the others. ConclusionsThese results are consistent with those found in groups of people with chronic pain taking larger doses of opioids and following opioid reduction or cessation programs. Opioid prescription did not increase over the 8 years, which was reassuring. These factors are important to emphasise because they can be modified in the rehabilitation setting with interdisciplinary management. RegistrationOur database was registered on Mendeley Data.