Improved detection tools are needed for the emerald ash borer, Agrilus planipennis Fairmaire (Coleoptera: Buprestidae), an invasive forest insect from Asia that has killed millions of ash (Fraxinus spp.) trees in North America since its discovery in Michigan in 2002. We evaluated attraction of adult A. planipennis to artificial traps incorporating visual (e.g., height, color, silhouette) and olfactory cues (e.g., host volatiles) at field sites in Michigan. We developed a double-decker trap consisting of a 3-m-tall polyvinyl pipe with two purple prisms attached near the top. In 2006, we compared A. planipennis attraction to double-decker traps baited with various combinations of manuka oil (containing sesquiterpenes present in ash bark), a blend of four ash leaf volatiles (leaf blend), and a rough texture to simulate bark. Significantly more A. planipennis were captured per trap when traps without the rough texture were baited with the leaf blend and manuka oil lures than on traps with texture and manuka oil but no leaf blend. In 2007, we also tested single prism traps set 1.5 m above ground and tower traps, similar to double-decker traps but 6 m tall. Double-decker traps baited with the leaf blend and manuka oil, with or without the addition of ash leaf and bark extracts, captured significantly more A. planipennis than similarly baited single prism traps, tower traps, or unbaited double-decker traps. A baited double-decker trap captured A. planipennis at a field site that was not previously known to be infested, representing the first detection event using artificial traps and lures. In 2008, we compared purple or green double-decker traps, single prisms suspended 3–5 m above ground in the ash canopy (canopy traps), and large flat purple traps (billboard traps). Significantly more A. planipennis were captured in purple versus green traps, baited traps versus unbaited traps, and double-decker versus canopy traps, whereas billboard traps were intermediate. At sites with very low A. planipennis densities, more A. planipennis were captured on baited double-decker traps than on other traps and a higher percentage of the baited double-decker traps captured beetles than any other trap design. In all 3 yr, peak A. planipennis activity occurred during late June to mid-July, corresponding to 800–1,200 growing degree-days base 10°C (DD₁₀). Nearly all (95%) beetles were captured by the end of July at ≈1400 DD₁₀.