In the present study, an ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) method was developed, validated, and applied for measuring cortisol in human hair. Baseline levels of cortisol in hair were taken from 12 control subjects, with concentrations for adult controls (n = 8) of 1.7 to 9.1 pg/mg and a median of 4.7 pg/mg and for child controls (n = 4) of 1.1 to 7.2 pg/mg and a median of 3.1 pg/mg. However, the concentrations in the hair of two children whose mother had been applying a cortisol-containing hand cream 2-3 times per week ranged from 30 to 390 pg/mg. No external contamination was observed with the children as judged from wash water concentrations. The mother had hair cortisol concentrations of 80-220 pg/mg. External contamination was observed in her proximal hair segments (0-4 cm) but not in distal ones (8-12 cm). In an experiment, cortisol cream (1%) was applied on the fingers of a subject, who then scratched the head hair once in a while. Hair was collected 1, 5, and 30 days after exposure to the cream. The cortisol level in the hair one day after exposure was 20-186 times higher than the pre-exposure level. High levels in the wash fraction agreed with external contamination. Cortisol concentrations in the hair at 5 and 30 days after exposure were 15-38 and 9-11 times higher, respectively, than the pre-exposure levels. However, no external contamination was suggested from the wash water concentrations in the hair collected 5 and 30 days after exposure. The results showed that the externally applied cortisol had, after some time, been incorporated into the hair matrix and was not removed by a pre-analysis washing. Therefore, the use of a standard decontamination procedure prior to analysis of hair may not be able to prevent the spread of cortisol from applied hand cream within a family.