The study characterizes the process of sensitization induced by intermittent administrations of quinpirole (0.5 mg/kg) in rats in a large open field. Sensitization was found to be self-limiting, with all measures of behavior reaching a plateau after the tenth twice-weekly injection. Kinetics of sensitization were a simple hyperbolic function of the number of drug injections for some measures (speed of locomotion, length of locomotor bouts) but showed positive co-operativity for others (distance travelled, duration of locomotion, frequency of stops, route stereotypy), suggesting potentiation of the effect by preceding injections. The pace of sensitization varied for different behaviors: locomotor speed changed fastest in the early portion of chronic treatment; stereotypy of route changed primarily during the late phase; mouthing did not sensitize. Sensitization evolved by a cascade of changes that included: advancing the onset of locomotor activation; prolonging the duration of locomotion; establishing new maxima of observable responses; altering the mode of locomotion; raising speed, rate and length of locomotor bouts; and increasing stereotypy of travel. These observations do not substantiate the prediction that development of behavioral sensitization is associated with emergence of disorganized activity and/or fractionation of response chains. Instead, it is proposed that development of sensitization may represent a build-up and strengthening of performance, reflecting enhanced central control of energy expenditure stimulated by repeated injections of quinpirole. Furthermore, it is suggested that for at least one response, the maximum observable amount of locomotion, development of sensitization requires only D2 stimulation, independent of D1 tone.