Emerald ash borer, Agrilus planipennis Fairmaire, an insect native to central Asia, was first detected in southeast Michigan in 2002, and has since killed millions of ash trees (Fraxinus spp.) in North America. The objectives of this study were to 1) determine an optimal sampling location on girdled ash trap trees for detection of A. planipennis larvae based on measurements of tree characteristics, and 2) develop a whole-tree estimation method for extrapolating larval densities from subsampled heights. We conducted sampling at 1-m increments, recording larval presence, height on tree bole, bolt diameter, and bark roughness for 58 infested ash trees. Analyzing height and diameter separately, generalized linear mixed models indicated the probability of A. planipennis detection was maximized at 17.2 cm for diameter and increased linearly as vertical height increased. There was also a positive relationship between intermediate bark roughness and A. planipennis presence. Stepwise regression indicated the optimal bolts for extrapolating whole tree larval densities were, in order of importance, at heights of 1–2 m, 4–5 m, 7–8 m, and 0–1 m. Subsampling with just one or two bolts explained 70% and 86%, respectively, of the variance in A. planipennis densities. Our results can be used by resource managers to improve efficiency of detection efforts and estimate infestations of A. planipennis.