Abstract The effect of cortisol on dreams has been scarcely studied. The aim of this exploratory study was to assess the possible effect of cortisol levels on dream recall/affect, considering, in female subjects, their menstrual cycle phase. Fifteen men and fifteen women were recruited. Saliva samples were used for the detection of cortisol levels. Participants were instructed to provide four saliva samples, during three consecutive days. After awakening, on the second and third day, they were asked whether they could recall the previous night’s dreams and whether these were pleasant or unpleasant. Female subjects followed this procedure twice: firstly, during the luteal phase and, secondly, during the follicular phase of the menstrual cycle. Subjects with higher evening or higher morning cortisol levels tended to show increased dream recall; a non-statistically significant association between morning cortisol levels and positive dream affect was also found. This association acquired statistical significance for salivary morning cortisol levels exceeding the upper normal level of 19.1 nmol/L (OR: 4.444, 95% CI: 1.108–17.830, p-value: 0.039). No connection between menstrual cycle stages and dream recall/affect was detected. In conclusion, cortisol may be a crucial neuromodulator, affecting dream recall and content. Therefore, its effects on sleep and dreams should be further studied.