This research investigated the relationship between the need for cognitive closure (NFC) and preference for generalized (over specialized) epistemic authorities. Seven studies tested the hypotheses that: individuals dispositionally higher (vs. lower) in NFC (1) evaluate generalized epistemic authorities more positively relative to specialized epistemic authorities, (2) report relying on generalized epistemic authorities more heavily relative to specialized epistemic authorities, and have epistemic authority sets that (3) are smaller and (4) consist of more generalized epistemic authorities; and individuals in whom NFC is situationally heightened (vs. lowered) (5) evaluate generalized epistemic authorities more positively relative to specialized epistemic authorities, (6) report a greater readiness to rely on more generalized (vs. specialized) epistemic authorities, and (7) report liking and relying more heavily on epistemic authorities framed as multifinal (vs. unifinal) means. Whereas the first six hypotheses outlined above concern the nature of the relationship between NFC and epistemic authority preferences, the seventh concerns the proposed mechanism, means multifinality, through which this link is established. The findings were mixed. Participants dispositionally higher in NFC did have epistemic authority sets consisting of more generalized epistemic authorities (Pilot Study A) and exhibited greater implicit liking of generalized epistemic authorities relative to specialized epistemic authorities (Study 1); however, the latter result was not obtained with explicit, self-report measures of liking (Studies 1, 2a,b, and 4). Moreover, unexpected results were obtained regarding NFC's relation to reliance on generalized (vs. specialized) epistemic authorities, with individuals higher in NFC, both dispositionally and situationally, exhibiting greater reliance on specialized (vs. generalized) epistemic authorities (Studies 1, 3, and 4). Experimental evidence from Study 4 suggests the means multifinality mechanism proposed to link NFC and epistemic authority generalization preferences is, in fact, in play, at least with respect to epistemic authority reliance; however, it appears to operate in a fashion opposite that predicted by the original theory. Finally, as predicted, NFC was inversely related to epistemic authority set size (Study 3 and Pilot Study B). A revised theory is presented to account for these findings, implications of the present research are discussed, and avenues for future research are suggested.