73 female university students were assigned to one of three groups, Internal, External, or No Traits, according to scores on Rotter's I-E Scale. Each group was divided into three conditions, Helpless, Non-helpless, and Control. During Phase I (treatment) subjects in the Helpless condition performed a functionally insoluble double-lever problem, subjects in the Non-helpless condition performed a similar but soluble problem, and Control subjects completed a noise-rating task. All subjects (Phase II) were then tested on a button-pushing problem which required both systematic and non-systematic strategies for solution. In the testing phase, subjects in the Helpless condition committed more errors and took more trials to criterion than did the Non-helpless and Control subjects. This difference between conditions was a result of poorer performance on the part of the Helpless subjects only during the phase of testing which required a systematic strategy. There were no main effects or interactions involving the locus of control variable on any of the dependent measures. The results were discussed in terms of previous research and an information-processing viewpoint was suggested as a possible alternative explanation of some learned helplessness phenomena.