Rats were trained to self-administer cocaine (0.5 mg/kg/infusion) and were then pretreated with the 5-HT1A agonist 8-OH-DPAT (0.125, 0.25 or 0.5 mg/kg, SC). 8-OH-DPAT pretreatment produced a decrease in reinforced response rates. When the effect of 8-OH-DPAT (0.5 mg/kg, SC) on responding for a range of cocaine doses was assessed, the drug produced a decrease in response rates when lower doses of cocaine served as the reinforcer. Fluoxetine (10 mg/kg, IV), an indirect 5-HT agonist, also reduced reinforced response rates for a low dose infusion of cocaine. Rates of responding for infusions of higher cocaine doses were not affected by fluoxetine pretreatment during an FR1 schedule of reinforcement. When an FR10 schedule of reinforcement was imposed, reinforced response rates for infusions of higher doses of cocaine were also reduced. Thus, under conditions that produce high rates of responding (low dose infusion or high ratio requirements for an infusion) fluoxetine reduced responding. This effect may be due to the effects at the 5-HT1A receptor, since 8-OH-DPAT produced a similar effect on cocaine self-administration. Given that the effects of these 5-HT agonists are observed only when low doses of cocaine serve as the reinforcer or when task demands are high, it is possible that the suppression of responding reflects an effect that is not specific to the reinforcing impact of cocaine. An alternative explanation for these effects incorporates a concept of unit cost/cocaine infusion that allows for direct comparison across studies that employ different reinforcement schedules.