We investigated reports of stunted growth and high mortality occurring among young poultry in 38 small privately owned flocks in Vermont and New Hampshire. Tests for infectious agents including avian influenza and Newcastle disease viruses were negative, as were toxicologic tests for mycotoxins, heavy metals, pesticides, and monensin in feed samples. Analysis of 6 samples of implicated feed revealed 330-870 ppm of sodium (median level 350 ppm), whereas the recommended minimum level for chick feed is 1200-2000 ppm. A case-control study included 38 affected and 23 unaffected flocks. Statistical analysis showed that affected flocks were 29 times more likely to have consumed a particular brand of poultry feed (odds ratio = 29.2, 95% confidence interval = 3.2-675.8) but did not show any association between clinical signs and hatchery, location of chick purchase, chick purchase date, or years of producer experience.