ABSTRACT: The relationships between infant feeding patterns, infant weight status, and subsequent growth are not yet firmly established. This study attempts to address these topics among Samoan children. The study population consists of 1,186 children born between October 1974 and September 197fi who were subsequently seen in the Well Baby Clinic in American Samoa. On the basis of the clinic records the children were categorized into feeding groups (breast vs. bottle) and weight-for-length groups (<25th percentile or >75th percentile of National Center for Health Statistics (NCHS) standards). A survey of 74 of these children was conducted in 1982. The average age of the children was 6.6 years. Twenty anthropometric measurements were taken and divided into three groups: skeletal, adipose, and mixed. Multiple analysis of variance was used on each of the groups of measurements to assess effects of infant feeding and weight-for-length. This analysis demon strafed significant relationships between infant weight-for-length and child hood skeletal size (heavy infants are larger children), between infant feeding patterns and childhood adiposity (bottle-fed children are fatter); and between both infant patterns and the mixed measurements (heavy and bottle-fed infants are heavier children with larger circumferences). The relationship between infant weight-for-length and skeletal size may be a continuation of infant patterns of high musculoskeletal weight in heavy infants. The relationship between infant feeding and childhood fatness was not affected by family income.