Optimal oviposition theory predicts that females prefer a host that assures the highest offspring fitness. I compared the host preference of females and larvae of a lichenivorous moth, Cleorodes lichenaria, and subsequent larval performance on host lichens with laboratory experiments and field observations. Both females and larvae preferred Ramalina fraxinea and Ramalina farinacea over Xanthoria parietina and Parmelia sulcata. Larvae reached the pupal stage faster on Ramalina species than on other lichens, whereas other fitness parameters including larval survival and pupal size did not differ between the lichen species. Interestingly, growth rates before overwintering were higher on P. sulcata and X. parietina than on Ramalina species, whereas after overwintering, the situation was reversed. The results of this study support the preference–performance hypothesis. Females clearly preferred host lichens that ensure the shortest developmental period at the expense of higher growth rates at the beginning of the larval period. Moreover, ecological and/or evolutionary factors in addition to females’ oviposition in host selection of C. lichenaria seem to be important. Because more larvae were found on Ramalina species than expected according to the oviposition pattern of females, final host selection is at least partially determined by larval behavior. It is also suggested that the physical and chemical properties of the hosts may provide a basis for enemy-free space for larvae.