To determine the effect of housing conditions on morphine self-administration, rats isolated in standard laboratory cages and rats living socially in a large open box (8.8 m2) were given morphine in solution (0.5 mg/ml) as their only source of fluid for 57 days. They were then exposed to a series of 3-day cycles previously shown by Nichols et al. (1956) to increase self-administration of morphine in caged rats. On morphine/water choice days late in the period of forced consumption, between the Nichols cycles, and during a subsequent period of abstinence, the isolated rats drank significantly more morphine solution than the social rats, and the females drank significantly more morphine solution than the males. During the four choice days in the Nichols Cycle Period the isolated rats increased their consumption, but the socially housed animals decreased theirs.