In an effort to assess models of inductive reasoning and deductive reasoning, the authors, in 3 experiments, examined the effects of argument length and logical validity on evaluation of arguments. In Experiments 1a and 1b, participants were given either induction or deduction instructions for a common set of stimuli. Two distinct effects were observed: Induction judgments were more affected by argument length, and deduction judgments were more affected by validity. In Experiment 2, fluency was manipulated by displaying the materials in a low-contrast font, leading to increased sensitivity to logical validity. Several variants of 1-process and 2-process models of reasoning were assessed against the results. A 1-process model that assumed the same scale of argument strength underlies induction and deduction was not successful. A 2-process model that assumed separate, continuous informational dimensions of apparent deductive validity and associative strength gave the more successful account. (Contains 2 tables and 9 figures.)