Recent preclinical and clinical data suggest that low dose of caffeine enhances the effects of common antidepressants. Here we investigated the effects of chronic administration of caffeine (5mg/kg, twice daily for 14days) and its withdrawal on day 15th on the activity of per se ineffective doses of fluoxetine (5mg/kg) and escitalopram (2mg/kg) given on day 15th. We found decreased immobility time in the forced swim and tail suspension tests in mice in which caffeine was administered simultaneously with antidepressants on day 15th following a 14-day caffeine treatment and no alterations in the spontaneous locomotor activity. A decrease in the level of escitalopram and an increase in the level of caffeine in serum were observed after concomitant administration of these compounds, while the joint administration of caffeine and fluoxetine was not associated with changes in their levels in serum or brain. Caffeine withdrawal caused a decrease in Adora1 mRNA level in the cerebral cortex (Cx). Administration of escitalopram or fluoxetine followed by caffeine withdrawal caused an increase in this gene expression, whereas administration of escitalopram, but not fluoxetine, on day 15th together with caffeine caused a decrease in Adora1 mRNA level in the Cx. Furthermore, antidepressant-like activity observed after joint administration of the tested drugs with caffeine was associated with decreased Slc6a15 mRNA level in the Cx. The results show that withdrawal of caffeine after its chronic intake may change activity of antidepressants with concomitant alterations within monoamine, adenosine and glutamate systems.