In the development of drugs that affect the central nervous system (CNS), it is important to determine whether they have dependence potential and if so, to establish an effective dose within a range that does not manifest dependency. Methods for evaluating drug dependence include self-administration tests using experimental animals while assessment of the minimum effective dose (MED) involves gross behavioral observation. This study aims to examine the relationship between the most frequently self-administered dose (peak self-administered dose, PSAD) and the MED amongst different types of CNS depressants in order to optimize dose range selection for the evaluation of drug reinforcing effects. PSAD was investigated by intravenous self-administration in rhesus monkeys conducted as daily 2 hr-sessions under a fixed ratio 5 schedule with a 1-min time-out after each administration. MED was investigated by gross behavioral observation following cumulative dosing. For opiates, the PSAD was 0.016 mg/kg/infusion for morphine, 0.06 mg/kg/infusion for codeine, 0.25 mg/kg/infusion for butorphanol and 0.063 mg/kg/infusion for pentazocine. For anesthetics and sedatives, the PSAD was 0.5 mg/kg/infusion for pentobarbital, 0.25 mg/kg/infusion for thiopental, 0.06 mg/kg/infusion for ketamine and 0.063 mg/kg/infusion for midazolam. The PSAD/MED ratio was 1/63-1/32 for opiates and 1/8-1/2 for anesthetics and hypnosedatives. While previous research by Fujiwara et al. (2016) suggested that a dose range lower than the MED for gross behavioral effects should be used for intravenous self-administration in the evaluation of drug reinforcing effects, this study further indicates that the optimal dose range may vary depending on drug type.