The incidence of primary urinary tract infection (UTI) is greatest in the first month of life and decreases with age throughout childhood. Secretory immunoglobulin A (sIgA) is an important component of mucosal immunity. The changes in secretory IgA, IgA and free secretory component (FSC) during the first year of life were examined in relation to age, sex and in infants, feeding practice. These constituents were further compared between healthy children and those with acute and recurrent UTI. Urine was collected from 41 healthy infants (16 female: 25 male) at intervals (mean age 1.4, 9.1, 44, 91, 210 and 412 days), 139 healthy children (75 female: 64 male), 29 children with histories of recurrent UTI (25 female: 4 male) and 10 with acute UTI (8 female: 2 male). sIgA, IgA and FSC were measured by enzyme linked immunoassay. In the majority of children sIgA and IgA were undetectable at birth. SIgA and IgA rose significantly during the first year then levelled off throughout childhood. FSC was detectable from birth (geometric mean [mean of logged values]-GOM at day 1.4, 362.2 ng/ml). No sex differences were apparent for any of the three constituents at any age. Breast feeding was associated with higher levels of sIgA and IgA than bottle feeding. This was highly significant at 9.1 days when sIgA and IgA levels of breast fed compared with bottle fed infants were 64.6 ng/ml vs 21.2 and 56.2 ng/ml vs 18.7 ng/ml respectively, giving a GOM ratio of 3.04 for sIgA and 3.0 for IgA (p < 0.001 for both). No significant difference in the three parameters were demonstrable when children with recurrent UTI-with normal or abnormal renal tracts-were compared with controls. Acute UTI resulted in raised sIgA, IgA and FSC compared with controls (GOM ratio of 4.9 [p < 0.002], 4.2 [p < 0.005] and 2.7 [p < 0.001] respectively). The proportion of total IgA present as sIgA (sIgA/total IgA) was not significantly different in the acute vs control groups. Urinary sIgA and IgA may be important for the observed variation with age in infant UTI and the reduced incidence in breast fed infants but does not appear to contribute to the sex associated difference in susceptibility to infection at any age.