The use in clinical practice and in clinical trials of the results of pain intensity measures is open to different views and interpretations. Clinically significant endpoints are seldom discussed. This study presents the use of three different outcome measures using 0 to 10 numerical scales to assess pain intensity in cancer patients in a controlled clinical trial on gabapentin and in a open label prospective study on PCA. Pain control was defined as daily pain score <5/10 (PI<5), or alternatively, pain intensity difference (PID) from baseline >/=2, or PID>/=33% of baseline pain intensity. The percentage of follow up days on which daily pain was controlled according to the 3 different criteria was chosen as individual longitudinal summary measure. Comparing the results of a randomized trial on gabapentin administration versus placebo for ten days a different behaviour could be shown for the three outcomes on a 10-day follow up period. Patients on gabapentin reached 100% of pain control in 8%, 23%, and 9% (versus placebo pts 7%, 15% and 7%) depending on the outcome measure adopted. Patients using morphine PCA for movement related pain for 3 to 7 days reached 100% pain control in 23% of cases (PI<5), 52% (PID>/=2) or 28% (PID>/=33%) of cases showing a significant difference in comparison with the control of pain at rest: PI<5 in 40%, PID>/=2 in 50%, and PID>/=33% in 35% of cases respectively. More complete analyses will be presented. The choice of the outcome measure could therefore change the sensitivity of the analysis performed and in it should be evaluated in relationship with the patients perception of pain control.