We used field surveys in central New Brunswick, Canada to establish efficient sampling procedures for evaluating densities of balsam gall midge, Paradiplosis tumifex Gagné (Diptera: Cecidomyiidae), and its associated damage in balsam fir, Abies balsamea (Linnaeus) Miller, Christmas trees. Infestation was greater in larger trees than smaller trees and in mid-crown and upper-crown branches than in the lower crown. However, the relationship between gallmaker infestation and site, height class, and crown level was highly complex and may involve covariation of shoot length with height class and crown level. As a result, patterns in infestation did not lend themselves to simple interpretation. This complexity highlights the need to find sampling units that provide simpler but reasonably accurate predictors of gallmaker impact at the whole-tree scale. We identified such a sampling unit: gallmaker density in first-order current-year shoots of a mid-crown branch explained 81% of the variance in total infestation among trees.