In 8 experiments using language processing tasks ranging from lexical decision to sensibility judgment, participants made hand or foot responses after reading hand- or foot-associated words such as action verbs. In general, response time (RT) tended to be faster when the hand- versus foot-associated word was compatible with the limb that was required to respond (e.g., hand response to a hand-associated word) than when it was incompatible (e.g., foot response to a hand-associated word). To see whether this compatibility effect reflects differential hand- versus foot-specific motor activation produced by the words, as suggested by some embodied theories of language understanding, we monitored 2 event-related potential (ERP) measures previously found to be sensitive to the activation of these limbs. As expected, the ERP results replicated previous findings that the monitored ERPs differ for hand versus foot movements. More importantly, the ERPs provided no evidence of any difference for hand- versus foot-associated words. Thus, the results weaken previous claims that the understanding of action verbs "requires" activation of the motor areas used to carry out the named action. Instead, they support claims that language-related compatibility effects on RT may arise prior to motor processes, which implies that such effects are not decisive evidence for embodied language understanding.