All communications on the use of transdermal fentanyl as well as the recommendations of the manufacturer include the direction that patients should be titrated with a short-acting narcotic to control their cancer pain before they are converted to a fentanyl transdermal therapeutic system (TTS). We investigated the possibility of avoiding this titration phase by immediate fentanyl TTS therapy in patients with uncontrolled cancer pain. Dose finding was performed by direct titration of fentanyl TTS according to clinical necessity on a day-to-day basis. Morphine solution for rescue medication was available. Short-term follow-up was 28 days, and 20 patients (10 in- and 10 outpatients) were evaluable. On the average, sufficient pain control [visual analogue scale (VAS) < 35 mm] was reached within 48 h of the start of fentanyl TTS. The mean VAS values before and during fentanyl TTS therapy were 53 mm (before), 27 mm (week 1), 21 mm (week 2), 18 mm (week 3) and 21 mm (week 4). There were statistically significant lower VAS values at all follow-up times compared to pretreatment values (e.g. pretreatment to day 1: P = 0.019; pretreatment to day 28: P = 0.002; Wilcoxon sign-rank test). The mean fentanyl TTS doses were 70 mug/h (week 1), 98 mug/h (week 2), 107 mug/h (week 3) and 116 mug/h (week 4). The differences of mean fentanyl TTS doses were significantly different between days 1 and 7 (P < 0.001) and between days 8 and 14 (P = 0.006), but not between days 15 and 21 and days 22 and 28. Mean morphine doses as rescue medication were steadily decreasing from 11 mg/day in week 1 to 3 mg/day in week 4 of treatment, but no statistically significant differences between these amounts could be found. Our results indicate that the titration phase with a short-acting narcotic prior to the conversion to fentanyl TTS is not necessary. Fentanyl TTS can be titrated safely and effectively on a day-to-day basis according to clinical necessity if the patients are well monitored, thus simplifying pain therapy with fentanyl TTS.