This study examines if the blocking effect paradigm predicts causal judgments when consequences of events vary in valence and magnitude. The procedure consists on presenting participants with reports describing the positive or negative effects produced by different substances, when these are consumed either separately or simultaneously with others. Two groups of participants were exposed to high and low magnitude consequences, respectively. The extent to which behavior with respect to causal judgments is consistent with the predictions of the blocking effect was evaluated in both groups using two types of questions. One of them asked whether or not substance X produced the effect, while the other one asked about the probability of substance X producing the effect. Differences in causal judgments as a product of logical or intuitive reasoning were examined. Even though the blocking effect was not observed, a significant interaction was obtained between the factors valence and experimental condition (blocking and control). Findings are discussed in terms of the differences between associative learning in humans and in non-human animals, and in terms of the theoretical differences between evaluative conditioning and predictive or causal conditioning.