SYNOPSISThis study aimed at comparing the time courses of pain relief in tension headache after taking 648 mg solid ASA and after taking 648 mg effervescent ASA (acetyl salicylic acid). A modified double dummy technique with both solid and effervescent placebo was used. 47 patients entered, and 33 patients took all four trial drugs. No statistically significant difference between solid and effervescent ASA could be demonstrated. ASA was significantly better than placebo although the median rating of ASA in this study was unexpectedly poor -as was the placebo response. This fact might have reflected the severely afflicted nature of a headache clinic's population.Different pain rating scales are compared, and it is concluded that the overall (GLOBAL) rating is sensitive, but should not be used alone, since it may sometimes reflect less relevant factors. The power of the trial was good, but reflects the comparatively poor responsiveness to ASA (low variance) in this particular patient population. An advantage of the effervescent preparation in more ASA-responsive patients cannot be ruled out. It is advised for future studies to recruit patients from general practice rather than from a headache-clinic population. (Headache 27:90-95, 1987) predict whether a difference could be shown in more responsive patients. The comparison of ASA Effervescent and ASA Tablets (both placebo corrected) was made at 30 minutes, as the difference between the plasma curves is known to be greatest at that time. It may take time, however, before the analgesic effect develops, and a later time would then have been more appropriate.Patient compliance. Lack of cooperation is not unknown in headache research. [26][27][28][29] Pharmacy students 17 or employees 2,3 seem more compliant than patients from a headache clinic. 28 In the light of our own previous good experience with migraine patients, 30 the poor compliance in the present study was disappointing and surprising. We can offer the following explanations: 1) Patients were rather severely afflicted and had no confidence in the test medication, which was explained as similar to aspirin. This was reflected in the very low placebo response. 2) Patients had to take 8 different doses. 3) the dose was probably suboptimal, and half of the doses were placebo. 4) Patients with tension headache are not bothered as much as migraine patients and may not find it worthwhile to spend the necessary time filling out forms etc. 5) Patients were instructed to take the medication for headaches of medium or severe intensity. This may partly explain why they took the test medication at relatively long intervals.In future trials patients should take test medications consecutively. To find responsive patients, recruitment from general practice seems advisable. It may be necessary to provide economic or other compensation to motivate such moderately afflicted patients. The test medication should be described realistically, but as positively as possible.