Introduction: The use of transdermal fentanyl is gaining in importance in the management of cancer pain. We describe the reasons for switching opioid medication to transdermal fentanyl in a pain management unit. Methods: Case records of patients treated with transdermal fentanyl in our pain clinic were evaluated retrospectively. Conversion ratios were calculated from the opioid dosage before and after conversion. Pain intensities were assessed on a numeric rating scale (NRS 0: no pain, 10: worst pain imaginable). Results: From October 1995 to December 1997 101 patients received transdermal fentanyl. Thirty-six patients had been treated with transdermal fentanyl before admission to our pain clinic, and relevant information was missing for one patient, so 64 patients were evaluated. Opioid therapy was switched to transdermal fentanyl during in-patient treatment for 53 patients and during outpatient treatment for 11 patients. Before conversion patients were treated with slow-release morphine (48%), immediate-release morphine (17%), buprenorphine (11%), tramadol (11%), levomethadone (5%), tilidine/naloxone (5%) and piritramid (3%). Reasons for opioid rotation were inadequate pain relief (33%), the patients' wish to reduce oral medication (20%), gastrointestinal side effects such as nausea (31%), vomiting (13%) and constipation (19%), dysphagia (27%) or others. Reduction of side effects was reported by 10 of 19 patients. In 12 of 21 patients, in whom the medication was switched because of inadequate pain relief, a reduction in pain intensity was reported. Discussion: Conversion to transdermal therapy may readjust the balance between opioid analgesia and side effects. The opioid switch resulted in more pain relief or fewer side effects in half of the patients. A proposed equianalgesic conversion ratio between 70:1 and 100:1 from oral slow-release morphine to transdermal fentanyl can be confirmed by our data. Conversion rates from other opioids to transdermal fentanyl are suggested.