OBJECTIVE: Humans experience the subjective effects of mu and kappa opioid agonists differently: mu agonists produce mainly euphoria, while kappa agonists are more likely to produce dysphoria. This study tested the hypothesis that these subjective effects would be associated with anatomically distinct changes in regional cerebral blood flow (CBF) relative to baseline as assessed with single photon emission computed tomography (SPECT). METHOD: Nine nondependent opioid abusers participated in the study. In the first phase of the study, the participants were acclimated to effects of the study drugs. In the second phase they underwent repeat challenges with the study drugs followed by an assessment of CBF with use of the SPECT tracer [99mTc]HMPAO. Medications tested were the prototypic mu agonist hydromorphone, the mixed agonist/antagonist butorphanol (which has a kappa agonist component of activity), and saline placebo. RESULTS: Subjective effects of the drugs were distinctly different. Hydromorphone produced increased ratings of "good effects," while butorphanol led to more "bad effects." Hydromorphone significantly increased regional CBF in the anterior cingulate cortex, both amygdalae, and the thalamus—all structures belonging to the limbic system. Butorphanol caused a less distinct picture of regional CBF increases, mainly in the area of both temporal lobes. CONCLUSIONS: This study demonstrates that opioids with different subjective effects also produce statistically significant patterns of change in regional CBF from baseline, and the regions of statistical significance appear in different brain regions. In addition, these results demonstrate the applicability of SPECT functional neuroimaging in the study of medications with potential abuse liability.