Project Summary Unpredictable drug access may play an important role in drug abuse. We know from laboratory studies that behavior maintained by nondrugs such as food is more persistent with unpredictable (i.e., variable) schedules of availability than with predictable (i.e., fixed) schedules, and nonhuman subjects choose a variable option over a fixed one even when it costs more in terms of responses per delivery. We recently found that rhesus monkeys choose variable cocaine over a fixed cocaine option, extending work with nondrug reinforcers to a drug of abuse. In addition, we have preliminary data showing that a history of responding for cocaine injections under a variable schedule increases cocaine self-administration under a progressive-ratio schedule. Thus, we predict that unpredictable access to drugs of abuse is an important determinant of the choice to take a drug at the expense of nondrug alternatives and that a history of unpredictable drug access can in- crease the reinforcing effectiveness of drugs. In Specific Aim 1, we will use drug vs. nondrug choice to evaluate the prediction that cocaine choice will be greatest when it is delivered under an unpredictable sched- ule and food under a predictable schedule. Conversely, cocaine choice will be decreased when food is deliv- ered under an unpredictable schedule and cocaine under a predictable schedule. Confirmation of our hypothe- ses would indicate that unpredictable availability contributes to excessive allocation of behavior toward procur- ing drugs at the expense of more predictable, nondrug alternatives. If unpredictable food delivery decreases cocaine choice, it would support the use of novel and easily implemented modifications to therapies like con- tingency management or treatments that incorporate nondrug reinforcers. In Specific Aim 2 we hypothesize that long-term exposure to a variable schedule of cocaine delivery will increase the effectiveness of cocaine as a reinforcer relative to a period of fixed cocaine exposure, and this effect will be enhanced with a more unpre- dictable VR compared to a more predictable VR. Confirmation of this prediction would indicate that unpredicta- bility is a contributing factor to the perseverative drug-seeking behavior seen in cocaine-dependent individuals. The data collected in this aim will allow us to evaluate whether placing drug reinforcers under more predictable schedules (as is experienced with agonist replacement therapy) can reduce some of the deleterious effects of a history of unpredictable access. A reduction in adverse consequences engendered by a switch from unpre- dictable to predictable schedules may represent a behavioral mechanism underlying agonist replacement ther- apies. Conversely, we will be able to evaluate wheth...