Aim-To establish a reference range for serum cortisol concentrations in preterm infants with a gestational age of less than 30 weeks during the first two weeks of life. Methods-Infants were prospectively classified by the following exclusion criteria: surfactant administration, arterial hypotension, acute or uncontrolled infection, ventricular haemorrhage II degrees or above, serum glucose < 2.2 mmol/l, exchange transfusion, stress as a result of any kind of examination or nursing for at least 4 hours before blood sampling. The cortisol value was measured once using radioimmunoassay in each infant. Results-In appropriate for gestational age (AGA) infants (n = 37, median gestational age 27.7 weeks, median birthweight 1030 g) the distribution of the cortisol concentrations was non-Gaussian. These had a nearly normal distribution, when log(10) values of the data were used. The points determined by mean (2 SD) on the logarithmic scale were transformed back to the original units to provide a reference range: 73-562 nmovl. Gestational age was significantly (p = 0.033) associated with cortisol values (log(10)) with a regression coefficient (standard error) of -0.045 (0.020). Small for gestational age (SGA) infants (n = 8) had significantly higher cortisol values (median 357 nmol/l) than AGA infants (median 199 nmol/l) (p = 0.028). Conclusions-There is a strictly defined reference range of serum cortisol concentrations in AGA preterm infants.