Purpose: The purpose of this exploratory study was to find a specific time of day with a stable cortisol level and to investigate the relationship between salivary cortisol and sleep. Methods: A total of 36 Korean female college students participated in the study. Salivary specimens were collected 6 times a day for 2 days in different stressful situations. Sleep characteristics were measured using an actigraph while salivary specimens were collected. Perceived stress was evaluated using the Global Assessment of Recent Stress. Results: Depending on whether there were morning peak and/or afternoon elevations in the cortisol levels, the type of diurnal cortisol pattern was classified into 4 types. None of the cortisol levels in different times of the day showed significant relationships to perceived stress levels. Cortisol levels in the morning, levels of peak cortisol and diurnal differences of cortisol were significantly correlated with sleep duration. The time with most stable cortisol level was 9-10 pm. Conclusion: It is recommended that measurements of salivary cortisol are taken from 9-10 pm since it showed a stable value regardless of diurnal cortisol rhythm and sleep. Sleep duration should be considered as an important confounding factor in measuring cortisol levels in the morning and the diurnal differences of cortisol.