Children who are abused have been said to have more illnesses than children who are not maltreated. The relationship between abuse and illness has been hypothesized to function in 2 ways: (1) that abuse precedes the illnesses and children from abusive homes become ill because of the damaging environment they endure, or, conversely, (2) that the illnesses precede the abuse, with the fussy behavior of ill children eliciting abuse. This study was intended to clarify the temporal relationship between illnesses and maltreatment. Health data were collected on a sample of 80 children: 11 from abusive families, 31 with nonorganic failure to thrive (NOFT), 14 from neglectful families, and 24 from control families. Hospital records (both inpatient and outpatient) for these children from the time of birth until they were 3 years old were searched by data collectors unaware of the child's classification. Children from abusive families or with NOFT appeared to be ill more often than control children, particularly during the first few months after birth, before abuse had been reported, but not necessarily before NOFT had been discovered. Health records of neglected children were not significantly different from those of controls. In addition to the abused. These 6 children also had more illnesses than control children, again particularly during the first few months after birth. Having ill children is described as a source of stress that may trigger abuse in an already stressed family.