We studied imidacloprid application methods and timing to control the hemlock woolly adelgid, Adelges tsugae Annand (Hemiptera: Adelgidae), in forests. The methods compared were 1) soil injection near the trunk; 2) soil injection dispersed throughout the area under the canopy; 3) soil drench near the base of the trunk; and trunk injection with the 4) Arborjet, 5) Wedgle, and 6) Mauget systems. The applications were made in the fall and the following spring. Adelgid populations on the hemlocks (Tsuga spp.) were assessed in the fall of two successive years after the treatments. Relative to the untreated control trees, all the soil applications resulted in population reductions, but none of the trunk injections resulted in reductions. Fall and spring treatment efficacy did not differ. Reductions by the soil treatments were between 50 and 100% (avg 80%) by the first fall and 83-100% (avg 98.5%) by the second fall. Analysis of imidacloprid residues using enzyme-linked immunosorbent assay found residues in sap, needles, and twigs 1 mo to 3-yr after application. A laboratory dose-response bioassay using excised, adelgid-infested hemlock branches with cut ends immersed in serial dilutions of imidacloprid determined the LC50 value to be 300 ppb, based on an exposure of 20 d. A high degree of suppression of the adelgid on forest trees was associated with residues in hemlock tissue > 120 ppb 2 yr after soil treatment. Although precise relationships between residues and efficacy are elusive, it is clear that soil application of imidacloprid resulted in chronic residues of imidacloprid in tissues and suppression of adelgid populations for > 2 yr.