In addition to providing pain relief, patient-controlled-analgesia (PCA) is also extensively used in clinical research for the assay of analgesic effectiveness of new drugs and methods of pain treatment. The main outcome measure of PCA analgesimetry is the difference in opioid requirements between the control (placebo) group and the new drug (or treatment) group. The following potential problems of PCA analgesimetry are analyzed: 1) weak correlation between pain intensity and opioid consumption, 2) interference of nonanalgesic effects of opioids, 3) role of acute tolerance to the analgesic effect of opioids, 4) problems of the patient's training, 5) interaction between main outcome measures, and 6) sample size and negative outcome problems. Knowledge of the pitfalls of PCA analgesimetry should decrease the risk of errors in its use.