Attention-deficit/hyperactivity disorder (ADHD) is comorbid with cocaine abuse. Whereas initiating ADHD medication in childhood does not alter later cocaine abuse risk, initiating medication during adolescence may increase risk. Preclinical work in the Spontaneously Hypertensive Rat (SHR) model of ADHD found that adolescent methylphenidate increased cocaine self-administration in adulthood, suggesting a need to identify alternatively efficacious medications for teens with ADHD. We examined effects of adolescent d-amphetamine treatment on strategy set shifting performance during adolescence and on cocaine self-administration and reinstatement of cocaine-seeking behavior (cue reactivity) during adulthood in male SHR, Wistar-Kyoto (inbred control), and Wistar (outbred control) rats. During the set shift phase, adolescent SHR needed more trials and had a longer latency to reach criterion, made more regressive errors and trial omissions, and exhibited slower and more variable lever press reaction times. d-Amphetamine improved performance only in SHR by increasing choice accuracy and decreasing errors and latency to criterion. In adulthood, SHR self-administered more cocaine, made more cocaine-seeking responses, and took longer to extinguish lever responding than control strains. Adolescent d-amphetamine did not alter cocaine self-administration in adult rats of any strain, but reduced cocaine seeking during the first of seven reinstatement test sessions in adult SHR. These findings highlight utility of SHR in modeling cognitive dysfunction and comorbid cocaine abuse in ADHD. Unlike methylphenidate, d-amphetamine improved several aspects of flexible learning in adolescent SHR and did not increase cocaine intake or cue reactivity in adult SHR. Thus, adolescent d-amphetamine was superior to methylphenidate in this ADHD model.