A reaction time study with normal human subjects was conducted to obtain support for one or the other model of interhemispheric relations - referred to as the one-system and the two-systems hypotheses. 12 subjects were extensively trained in a complex reaction time task consisting of a "priming sub-task" which introduced interhemispheric interference and two different "test sub-tasks" measuring the generalization of interference. The interference priming produced visible slowing of RT's on subsequent trials. Interestingly, the deterioration of the two test sub-tasks was alike, despite the marked difference in the amount of the interhemispheric communication they required. This result is more in line with the one-system hypothesis, as the two-systems hypothesis predicts deterioration proportional to the amount of the required interhemispheric communication.