Avian populations in Hanover, N. H., a town that has sprayed its elms with DDT for many years in an attempt to control Dutch elm disease, were compared with those in Norwich, Vt., a town 1 mile (1.6 km) west of Hanover that has never sprayed. Hanover applied 109 lb DDT/acre (2.1 kg/hectare) in April 1963, then used Methoxychlor in April 1964. Population surveys were taken regularly during spring and early summer of these years, dead birds were collected in both towns, and 106 birds were analyzed for DDT, DDE, and DDD. Severe mortality of both resident and migrant birds occurred in Hanover during spring 1963, and the evidence implicates DDT as its cause. Robin loss was estimated at 70% of the resident population, or 350 to 400 individuals, but mortality among other species of widely varied feeding habits was also substantial. Feeding habits suggest that some birds acquired the toxicant by eating living insects carrying DDT, presenting the paradox of survival of the intended DDT victims, and death, instead, of insectivorous birds. Organ and whole bird analyses are presented and criteria for establishing cause of death are discussed. Most of the DDT had been converted to DDE and DDD, and residues were found in all organs analyzed. Robin mortality was reduced, but not eliminated following Methoxychlor application in 1964; these losses were believed caused by residual DDT in the soil. There was no evidence DDT poisoning among other species in 1964, though the dead birds collected were not analyzed.