The HIV‐1 transgenic (Tg) rat model is valuable for understanding HIV‐associated neurocognitive disorders (HAND) and accompanying substance use and misuse. Tg and F344/NHsd wildtype (WT) rats were allowed to self‐administer intrajugular cocaine. For the first 7 sessions, neither genotype self‐administered cocaine (0.1 mg/kg/infusion) on a fixed ratio 1 schedule. We thus implemented a lever–cocaine “autoshaping” session followed by a series of manipulations changing dose and reinforcement schedule. Tg rats self‐administered much less cocaine than WT rats throughout the study. Of 8 Tg rats, 5 modestly increased self‐administration from sessions 36–50. Of those, only 3 showed a lever discrimination. Of 10 WT rats, 8 acquired robust self‐administration by session 19; all WT rats self‐administered cocaine by the end of the study. WT and Tg rats had similar baseline locomotor activity in the self‐administration chamber suggesting that the low levels of cocaine intake in the Tg rats did not reflect a nonspecific motor impairment in this rat strain. Concomitant measurement of activity with self‐administration revealed activity increases that followed increased cocaine intake. That relation held in Tg rats. Therefore, the present study provides evidence that HIV‐1 Tg rats are less sensitive to the reinforcing effects of cocaine than their F344 WT counterparts.