Some controversy has surrounded the question whether the second language (L2) of bilinguals can access conceptual memory as directly as their first language (L1). It has been proposed that an early age of acquisition and a high degree of proficiency are necessary for direct conceptual access. In the present study, Swedish undergraduates, with English as their second language, and a relatively late (9 yrs) age of acquisition, performed two word processing tasks in both L1 and L2. Exp. 1 used a picture-name verification task, where pictures induced automatic semantic priming. Exp. 2 used a category-exemplar verification task, where pictures again induced semantic priming, although in this case of a strategic kind. In both Exp. 1 (n = 22) and Exp. 2 (n = 23), the primary measure of semantic priming was reduction of the N400 amplitude in the ERP. Exp. 1 (automatic priming) showed a fronto-central priming effect that was indistinguishable in L1 and L2, as to amplitude, topographical distribution, and peak latency, although onset was somewhat earlier in L1. The effects of priming on reaction time (RT) were also parallel in L1 and L2. Exp. 2 (strategic priming) showed a different, centro-parietal priming effect that was similar in amplitude between L1 and L2, but differed in peak latency, and lateral distribution between languages. Also, the language factor interacted with priming in its effects on RT. The study indicates that L2 provides direct automatic access to conceptual memory, although strategic use of it may recruit partly different neuronal resources in L2 than in L1.