Eriophyoid mites, which are among the smallest plant feeders, are characterized by the intimate relationships they have with their hosts and the restricted range of plants upon which they can reproduce. The knowledge of their true host ranges and mechanisms causing host specificity is fundamental to understanding mite-host interactions, potential mite-host coevolution, and diversity of this group, as well as to apply effective control strategies or to use them as effective biological control agents. The aim of this paper is to review current knowledge on host specificity and specialization in eriophyoid mites, and to point out knowledge gaps and doubts. Using available data on described species and recorded hosts we showed that: (1) 80% of eriophyoids have been reported on only one host species, 95% on one host genus, and 99% on one host family; (2) Diptilomiopidae has the highest proportion of monophagous species and Phytoptidae has the fewest; (3) non-monophagous eriophyoids show the tendency to infest closely related hosts; 4) vagrant eriophyoids have a higher proportion of monophagous species than refuge-seeking and refuge-inducing species; (5) the proportions of monophagous species infesting annual and perennial hosts are similar; however, many species infesting annual hosts have wider host ranges than those infesting perennial hosts; (6) the proportions of species that are monophagous infesting evergreen and deciduous plants are similar; (7) non-monophagous eriophyoid species have wider geographic distribution than monophagous species. Field and laboratory host-specificity tests for several eriophyoid species and their importance for biological control of weeds are described. Testing the actual host range of a given eriophyoid species, searching for ecological data, genetic differentiation analysis, and recognizing factors and mechanisms that contribute to host specificity of eriophyoid mites are suggested as future directions for research.