Background: Breakthrough pain (BTP) has not formerly been discussed as such in chronic non-malignant pain patients referred to pain centres and clinics. The purpose of the study was to investigate the prevalence, characteristics and mechanisms of BTP in opioid-treated chronic non-malignant pain patients referred to a pain centre and to assess the short-term effects of pain treatment. Methods: Patients were assessed at referral (T0) and after a treatment period of 3 months (T3) using the visual analogue scale (VAS) of the brief pain inventory (BPI) within somatic nociceptive, neuropathic and/or visceral pain conditions, the mini mental state examination (MMSE) and the hospital anxiety and depression scale (HADS). The main treatment intervention from T0 to T3 was to convert short-acting oral opioids to long-acting oral opioids and to discontinue on demand and parenteral use of opioids. Results: Thirty-three patients were assessed at T0 and 27 at T3. The prevalence of BTP declined significantly from T0 (90%) to T3 (70.4%). Worst, least, average and current pain intensities as well as duration of BTP were significantly reduced from T0 to T3. The majority of BTPs were exacerbation of background pain assumed to be of the same pain mechanisms. High average pain intensity (BPI) was significantly associated with high scores for both anxiety and depression (HADS). Conclusion: BTP in chronic non-malignant pain patients seems to be surprisingly frequent and severe. Stabilizing the opioid regimen seems to reduce pain intensity in general as well as the intensity and duration of BTP. Average pain intensity was associated with anxiety and depression.