Social buffering is the phenomenon in which the presence of an affiliative conspecific mitigates stress responses. We previously demonstrated that social buffering completely ameliorates conditioned fear responses in rats. The present study explored the neurochemical background of this social buffering. In Experiment 1, fear-conditioned subjects first received an intraperitoneal injection of either naloxone (non-selective opioid receptor antagonist), haloperidol (dopamine D2 receptor antagonist), SR49059 (vasopressin V1A receptor antagonist), atosiban (oxytocin receptor antagonist), or saline. The subjects were then exposed to a conditioned stimulus with an unfamiliar non-conditioned rat. Naloxone, but not the other three antagonists, blocked social buffering. In Experiment 2, we assessed the effect of naloxone on locomotor activity during an open-field test. Naloxone did not affect walking steps during the test. Therefore, it is unlikely that the results of Experiment 1 were due to decreased activity by naloxone. In Experiment 3, we assessed Fos expression in 16 brain regions accompanied by the blockade of social buffering by naloxone. Consistent with the results of Experiment 1, Fos expression was increased in the paraventricular nucleus of the hypothalamus. In addition, Fos expression was decreased in the nucleus accumbens shell, anterior cingulate cortex, and insular cortex and tended to be decreased in the nucleus accumbens core. Naloxone thus appears to affect these four regions and/or act upstream of these regions during blockade of social buffering. Based on these results, we conclude that naloxone blocks social buffering of conditioned fear responses in rats.