The 5 experiments in this article examine whether bihemispheric processing can be predicted from the processing observed when information is directed initially to just 1 hemisphere. Ss decided if laterally presented words rhymed with a previously presented central target. Trials varied in the degree to which the information displayed was redundant. On some trials, the same word appeared twice; in others, different words appeared, but they both led to the same decision; and finally, on some trials, different words appeared, each of which also led to a different decision. The patterns found for unilateral and bilateral trials were distinct. Furthermore, the pattern observed on bilateral trials could not be predicted from that found on unilateral trials, suggesting that interhemispheric interaction cannot be surmised from observing a hemisphere processing information in relative isolation.