Effective methods for early detection of newly established, low density emerald ash borer (Agrilus planipennis Fairmaire) infestations are critically needed in North America. We assessed adult A. planipennis captures on four types of traps in a 16-ha site in central Michigan. The site was divided into 16 blocks, each comprised of four 50- by 50-m cells. Green ash trees (Fraxinus pennsylvanica Marshall) were inventoried by diameter class and ash phloem area was estimated for each cell. One trap type was randomly assigned to each cell in each block. Because initial sampling showed that A. planipennis density was extremely low, infested ash logs were introduced into the center of the site. In total, 87 beetles were captured during the summer. Purple double-decker traps baited with a blend of ash leaf volatiles, Manuka oil, and ethanol captured 65% of all A. planipennis beetles. Similarly baited, green double-decker traps captured 18% of the beetles, whereas sticky bands on girdled trees captured 11% of the beetles. Purple traps baited with Manuka oil and suspended in the canopies of live ash trees captured only 5% of the beetles. At least one beetle was captured on 81% of the purple double-decker traps, 56% of the green double-decker traps, 42% of sticky bands, and 25% of the canopy traps. Abundance of ash phloem near traps had no effect on captures and trap location and sun exposure had only weak effects on captures. Twelve girdled and 29 nongirdled trees were felled and sampled in winter. Current-year larvae were present in 100% of the girdled trees and 72% of the nongirdled trees, but larval density was five times higher on girdled than nongirdled trees.