Superconditioning refers to supernormal responding to a conditioned stimulus (CS) that sometimes occurs in classical conditioning when the CS is paired with an unconditioned stimulus (US) in the presence of a conditioned inhibitor for that US. In the present research, we conducted 4 experiments to investigate causal superlearning, a phenomenon in human causal learning analogous to superconditioning. Experiment 1 demonstrated superlearning relative to appropriate control conditions. Experiment 2 showed that superlearning wanes when the number of cues used in an experiment is relatively large. Experiment 3 determined that even when relatively many cues are used, superlearning can be observed provided testing is conducted immediately after training, which is problematic for explanations by most contemporary learning theories. Experiment 4 found that ratings of a superlearning cue are weaker than those to the training excitor which gives basis to the conditioned inhibitor-like causal preventor used during causal superlearning training. This is inconsistent with the prediction by propositional reasoning accounts of causal cue competition, but is readily explained by associative learning models. In sum, the current experiments revealed some weaknesses of both the associative and propositional reasoning models with respect to causal superlearning.