Two experiments were performed to discriminate between a pathway inhibition and a response competition explanation of the negative effects of semantic priming on recall. Both experiments involved recalling a word in response to a definition which was preceded by various types of priming stimuli. Experiment 1 was designed to differentiate between these explanations by using category name and instance priming stimuli. In agreement with the response competition hypothesis, priming with an instance of the same category of the target increased the latency to recall that target and increased the number of erroneous responses. In contrast, priming with the category name of the target was found to either facilitate or have no effect on recall latency and decrease the number of errors. Thus, the results of experiment 1 clearly favored the response competition hypothesis of the negative effects of semantic priming on recall. Experiment 2 performed the same theoretical function by manipulating the number of semantic priming stimuli and the typicality of the priming instances in relation to the category of the correct target. However, experiment 2 did not present convincing evidence for or against either the response competition or pathway inhibition hypotheses. The results showed little or no effect of the number of priming stimuli or their semantic relatedness to the target. These results were tentatively explained by methodological considerations. Implications concerning response competition in semantic retrieval and problem solving are also discussed.