Lake trout (Salvelinus namaycush) is the presumed preferred host of the invasive sea lamprey (Petromyzon marinus) in the Great Lakes, but little is understood about this preference outside of laboratory experiments. By preference we mean sea lamprey attacks on hosts are disproportionate to host relative abundance. The purpose of this study was to quantify host preference of sea lampreys in the field for the first time. We focused our analysis on Lake Ontario, where the two dominant host species for sea lampreys are lake trout and Chinook salmon (Oncorhynchus tshawytscha). Sea lampreys exhibited a strong preference for lake trout when host abundance was ≥ 32% lake trout, but sea lamprey preference switched to Chinook salmon when host abundance was ≤ 13% lake trout. Model results were equivocal with respect to determining whether the primary predictor of sea lamprey preference was relative or absolute abundance of lake trout. Other evidence, particularly the different spatial distributions of the two hosts, suggests that sea lamprey preference for lake trout is based on a higher encounter probability, driven by absolute abundance of lake trout. This study confirms a widely held suspicion that observed sea lamprey-induced marking rates on lake trout, used to assess the status of sea lamprey control in the Great Lakes, can be influenced by the abundance of alternative hosts. As an alternative host for sea lamprey parasitism, Chinook salmon may contribute to the persistence of lake trout in Lake Ontario.