Recent evidence suggests that sleep deprivation leads to suboptimal decision-making on the Iowa Gambling Task (IGT), a pattern that appears to be unaffected by moderate doses of caffeine. It is not known whether impaired decision-making could be reversed by higher doses of caffeine or by other stimulant countermeasures, such as dextroamphetamine or modafinil. Fifty-four diurnally active healthy subjects completed alternate versions of the IGT at rested baseline, at 23 and 46 h awake, and following a night of recovery sleep. After 44 h awake, participants received a double-blind dose of caffeine (600 mg), dextroamphetamine (20 mg), modafinil (400 mg), or placebo. At baseline, participants showed a normal pattern of advantageous performance, whereas both sleep-deprived sessions were associated with suboptimal decision-making on the IGT. Following stimulant administration on the second night of sleep deprivation, groups receiving caffeine, dextroamphetamine, or modafinil showed significant reduction in subjective sleepiness and improvement in psychomotor vigilance, but decision-making on the IGT remained impaired for all stimulants and did not differ from placebo. Decision-making returned to normal following recovery sleep. These findings are consistent with prior research showing that sleep deprivation leads to suboptimal decision-making on some types of tasks, particularly those that rely heavily on emotion processing regions of the brain, such as the ventromedial prefrontal cortex. Moreover, the deficits in decision-making were not reversed by commonly used stimulant countermeasures, despite restoration of psychomotor vigilance and alertness. These three stimulants may restore some, but not all, aspects of cognitive functioning during sleep deprivation.