In a prospective follow-up study, 60 children exposed to alcohol in utero were assessed by a psychologist (Bayley Mental scale) and a speech therapist (Reynell Verbal Comprehension scale) at a mean age of 27 months. Many mothers had been able to reduce their alcohol consumption during pregnancy, so the children could be divided into those exposed to heavy drinking during the first trimester only (group 1, n = 20), those exposed during the first and second trimesters (group 2, n = 20), and those exposed throughout pregnancy (group 3, n = 20). Forty-eight nonexposed children were examined to set the -2 SD limit for subnormal performance on the Bayley and Reynell tests. No definite effect of alcohol exposure on mental or language development was found in group 1. Children in group 3 scored significantly lower than children in group 1 both on the Bayley Mental scale and on the Reynell Verbal Comprehension scale; delay in language development was seen more often in group 2 than in group 1. The diagnosis of fetal alcohol syndrome was made in seven children (one in group 2 and six in group 3) and the diagnosis of fetal alcohol effects in 13 children (one in group 1, three in group 2, and nine in group 3). Efforts should be made to identify and find proper treatment for women who drink alcohol early in their pregnancies.