Low birth weight is associated with increased cardiovascular and metabolic disorders in adult life, although the mechanisms of this effect remain uncertain. There is one report of increased morning plasma cortisol levels in an elderly low birth weight cohort, but whether this is primary or secondary to other aspects of the phenotype is unclear. We investigated the association between low birth weight and glucose intolerance, blood pressure, and dyslipidemia in young, nonobese adults from a community undergoing the health transition with a high prevalence of both noncommunicable diseases and low birth weight. Additionally, we investigated whether altered basal and stimulated cortisol levels as a marker of hypothalamic-pituitary-adrenal responsiveness or cortisol metabolism were associated with low birth weight in these young adults. Twenty-year-old, historically disadvantaged, urbanized South Africans (n = 137) with birth weights either below the 10th percentile [underweight for age (UFA)] or between the 25th and 75th percentiles [appropriate for gestational age (AFA)] had anthropometry, blood pressure, lipid levels, and glucose tolerance measured. In a subset (n = 62), 0900 h plasma cortisol concentrations, cortisol responses to 1 microg ACTH, and urinary glucocorticoid metabolites were measured. The mothers of UFA infants were themselves lighter and had a lower body mass index (P: = 0. 0016). At age 20 yr, although the UFA group was still smaller and lighter, with a lower body mass index, they had higher fasting plasma glucose levels (P: = 0.047), and a greater proportion demonstrated glucose intolerance (11.9% vs. 0%; P: < 0.01). The UFA group also had higher systolic [UFA, 126.0 +/- 13.3 (+/-SD); AFA, 122.0 +/- 11.7 mm Hg; P: = 0.007] and diastolic (72.3 +/- 8.4 vs. 69. 5 +/- 8.7 mm Hg; P: = 0.02) blood pressures, after covarying for current weight and gender. Plasma cortisol levels determined at 0900 h were higher in the UFA group (484.9 +/- 166.3 vs. 418.6 +/- 160.6 nmol/L) and showed a greater plasma cortisol response to low dose ACTH stimulation (area under the curve for cortisol: UFA, 77,238 +/- 19,511; AFA, 66,597 +/- 16,064 nmol/L.min; P: = 0.04). In conclusion, the link between low birth weight and adult glucose intolerance and blood pressure elevation occurs in young adults in a high risk, disadvantaged population despite a lack of full catch-up growth. Moreover, cortisol axis activation is an early feature in the process linking low birth weight with adult cardiovascular and metabolic disease and is not dependent upon adult obesity or full catch-up growth, at least in this population undergoing the health transition.