Understanding how genetic variation is maintained in ecologically-important traits is a fundamental question in evolutionary biology. Male Trinidadian guppies (Poecilia reticulata) exhibit extreme genetic diversity in color patterns within populations, which is believed to be promoted by a female mating preference for rare or novel patterns. However, the origins of this preference remain unclear. Here, we test the hypothesis that mating preference for novel phenotypes is a by-product of general neophilia that evolved in response to selection in non-mating contexts. We measured among-female variation in preference for eight different, novel stimuli that spanned four ecological contexts: mate choice, exploration, foraging, and social (but non-sexual) interactions. Females exhibited preference for novelty in 6 out of 8 tests. Individual variation in preference for novelty was positively correlated among all 8 types of stimuli. Furthermore, factor analysis revealed a single axis of general neophilia that accounts for 61% of individual variation in preference for novel color patterns. The single-factor structure of neophilia suggests that interest in novelty is governed primarily by shared processes that transcend context. Because neophilia likely has a sizable heritable component, our results provide evidence that mating preference for novel phenotypes may be a non-adaptive by-product of natural selection on neophilia.