This study explored the hypothesis that preschoolers' deductive reasoning would be improved by encouraging use of divergent thinking (DT). Children of 4-5 years of age (n = 120) were randomly given DT or neutral control exercises before deductive reasoning problems. To allow a stronger test of the hypothesis, half of the children receiving the DT exercises were given explicit examples, which have been shown to reduce ideational originality. Results indicate that, as predicted, DT exercises without examples significantly improved rates of deductive responding, compared to exercises with examples and the control condition. These findings indicate that DT is a key component in the early beginnings of deductive reasoning. Some educational implications are discussed.