Studies using rats in an alleyway are reviewed with regard to the partial reinforcement acquisition and extinction effects. The magnitude of these effects is complicated by the fact that certain variables have differential effects depending on whether the rat is given a partial or a continuous reinforcement schedule. Three major theoretical positions, dissonance, frustration, and habituation of competing responses, are shown to encounter difficulties with a considerable body of data. It is suggested that extensions and modifications of Capaldi's sequential theory, which handled most aspects of the extinction effect, may prove to be a viable theoretical alternative.