Emerald ash borer, Agrilus planipennis Fairmaire (Coleoptera: Buprestidae), has killed millions of ash (Fraxinus sp.) trees in North America since its discovery in Michigan in 2002. Efficient methods to detect low-density A. planipennis populations remain a critical priority for regulatory and resource management agencies. We compared the density of adult A. planipennis captured on sticky bands and larval density among ash trees that were girdled for 1 or 2 yr, wounded, exposed to the stress-elicitor methyl jasmonate, baited with Manuka oil lures, or left untreated. Studies were conducted at four sites in 2006 and 2007, where A. planipennis densities on untreated trees ranged from very low to moderate. In 2006, 1-yr girdled trees captured significantly more adult A. planipennis and had higher larval densities than untreated control trees or trees treated with methyl jasmonate or Manuka oil. Open-grown trees captured significantly more A. planipennis beetles than partially or fully shaded trees. In 2007, A. planipennis population levels and captures of adult A. planipennis were substantially higher than in 2006. The density of adults captured on sticky bands did not differ significantly among canopy exposure classes or treatments in 2007. Larval density was significantly higher in untreated, wounded, and 1-yr girdled trees (girdled in 2007) than in 2-yr girdled trees (girdled in 2006), where most phloem was consumed by A. planipennis larvae the previous year. A total of 36 trees (32 in 2006, 4 in 2007) caught no beetles, but 16 of those trees (13 in 2006, 3 in 2007) had A. planipennis larvae. In 2006, there was a positive linear relationship between the density of adults captured on sticky bands and larval density in trees. Our results show that freshly girdled and open grown trees were most attractive to A. planipennis, especially at low-density sites. If girdled trees are used for A. planipennis detection or survey, debarking trees to locate larval galleries is crucial.