Sherman and MItchell (1972) have described a drug-test interaction between morphine and pain which consisted of three phases: (1) a reduction in levels of Krebs cycle intermediaries by morphine, (2) a reduction due to pain, and (3) a return toward control levels with the combination of morphine and pain. Three studies were performed to determine (1) whether the drug-test interaction was also observed in mouse brain or liver, (2) the effect of tolerance on the development of this phenomenon, and (3) the effect of naloxone on this phenomenon. Male CF1 mice were given either saline or morphine and either exposed to analgesic testing or not, then frozen in liquid nitrogen. Brains and livers were analysed for acids of the citric acid cycle by gas chromatography. In experiment II, a tolerant population was used. In experiment III, the effect of naloxone was observed. In experiment I, the three-part interaction was observed in brain, but not in liver. Experiment II showed that tolerance prevented the occurrence. Experiment III showed that naloxone counteracted the effect of morphine on pain. However, naloxone itself also blocked the alteration in Krebs cycle intermediaries produced by pain. This latter finding argues against these changes in Krebs cycle intermediaries being related to either pain or analgesia but they may be related to some other action of morphine.