A retrospective review of the hospital records of New York City children aged 6 months through 6 years showed that 63 homeless children had a higher rate of immunization delay than an age- and sex-stratified sample of 63 domiciled children living at the same federal poverty level. In a logistic regression model, this difference persisted after controlling for sex, age, ethnicity, presence of chronic illness, and reason for referral. In a 6-month- to 2-year-old subgroup, homeless and domiciled children had equal rates of anemia, but homeless children were more likely to have elevated erythrocyte protoporphyrin (EP) levels consistent with iron deficiency. This difference, too, persisted after controlling for the same confounding factors. Elevated EP levels and immunization delay were likely to coexist in the homeless children. The higher rate of immunization delay is compatible with the occurrence of measles outbreaks in some New York City shelters. The higher rates of iron deficiency may reflect overall poor nutrition. All these findings have significant implications for the design of health-care programs for homeless children.