Objective Is the involvement of the regular general practitioner (GP) in the decision to initiate opioid treatment for chronic non-cancer pain (CNCP) associated with two main risk factors for serious adverse events: increased opioid dose and the concomitant use of prescribed benzodiazepines or benzodiazepine-related medications? Design and setting An anonymous web-based survey was conducted in the county of Rogaland, Norway, during the spring of 2021. Subjects GPs who self-reported applying at least once for reimbursement of opioids prescribed to treat CNCP. They were asked to answer the survey based on the last patient for whom they recalled submitting a reimbursement application. Main outcome measures 1) Total opioid dose in daily oral morphine equivalents (OMEQ). 2) Concurrent use of benzodiazepines and/or benzodiazepine-related drugs. Results The daily opioid dose was lower when the surveyed GPs initiated the opioid treatment (36 OMEQ, n = 25), than when others had initiated the treatment (108 OMEQ, n = 31, p = 0.001). For concurrent use of benzodiazepine or benzodiazepine-related drugs, no significant difference was found (33%, n = 9 with GP involvement vs. 47%, n = 16, p = 0.279 with no GP involvement). Conclusions GP involvement in the initiation of opioid medication for CNCP was associated with a lower opioid dose being prescribed. Implications GP involvement in the initiation of opioid prescriptions may facilitate safer prescribing.