Drug-naive dogs were trained to respond for intravenous infusions of either d-amphetamine, phenmetrazine, or methylphenidate until a stable response rate per 4-hr daily session was achieved. The magnitude of reinforcement (i.e., mg/kg/infusion) was then varied systematically across a wide range for each drug. An inverse relationship between unit dose and number of self-administered infusions per session was seen. Thus, total drug intake per session remained relatively constant and was independent of unit dose. Using a parallel line bioassay design, the relative potencies of d-amphetamine, phenmetrazine, and methylphenidate to maintain self-administration were estimated. By comparing the unit doses of d-amphetamine which yielded the same rate of self-administration it was found that 1 mg of phenmetrazine is equivalent to 0.1 mg of d-amphetamine. It was also determined that 1 mg of methylphenidate is equivalent to 0.75 mg of d-amphetamine. These data indicate the dog can be used to assess the reinforcing properties of psychomotor stimulants.