Several studies evidenced the importance of the knowledge of the bioclimatic comfort for improving people's quality of life. Temperature and relative humidity are the main variables related to climatic comfort/discomfort, influencing the environmental stress in the human body. In this study, a stochastic approach is proposed for characterizing the bioclimatic conditions through the Humidex values in six sites of Calabria (southern Italy), a region often hit by heat waves in summer months. The stochastic approach is essential, because the available time series of temperature and relative humidity are not long enough and present several missing values. The model allowed the characterization of sequences of extreme values of the Humidex. Results showed different behaviours between inner and coastal stations. For example, a sequence of 20 consecutive days with maximum daily Humidex values greater than 35 has a return period ranging from 10 to 20 years for the inner stations, while it exceeds 100 years for the coastal ones. The maximum yearly Humidex values for the inner stations have a larger range (40-50) than the coastal ones (38-45), reaching higher occurrence probabilities of serious danger conditions. Besides, the different influence of temperature and relative humidity on the Humidex behaviour has been evidenced.