BIA 10-2474 (a fatty acid amide hydrolase inhibitor) was evaluated in a first-in-human phase 1 study in normal volunteers to assess safety/tolerability, pharmacokinetics, pharmacodynamics, and food effect. The dose-escalation process consisted of a single-ascending-dose phase (SAD) and multiple-ascending-dose phase (MAD). Prospective determination of the starting dose and maximal escalated dose was consistent with the usual clinical pharmacology principles for extrapolation of preclinical toxicology data to human equivalent doses. After only 5-6 days of multiple-dose administration of 50 mg daily in the MAD phase, several subjects became quite ill with central nervous system symptoms. One subject progressed to brain death within several days of symptom onset. Magnetic resonance imaging scans demonstrated signal abnormalities consistent with microbleeds affecting the hippocampus and pons, suggestive of possible cytotoxic or vasogenic edema compatible with a toxic/metabolic process. There were no findings at lower MAD doses or during the SAD phase. The toxicology program carried out in 4 preclinical species (mouse, rat, dog, and monkey) did not demonstrate significant neurotoxicity. The probable mechanism of neurologic toxicity demonstrated in humans at the 50-mg daily dose was inhibition of off-target cerebral receptors or through another mechanism. Additional recommendations have been proposed for future first-in-human studies to maximize subject safety. However, one must also accept the basic premise that, in general, first-in-human phase 1 studies are remarkably safe, and these rare events are not 100% avoidable during the drug development process.