High intensity training regimens appear to put athletes at a higher risk of illness. As these have been linked to alterations in the proportions of differentiated T cells, how training load affects these populations could have important implications for athlete susceptibility to disease. This study examined the effect of a winter training season on the proportions of circulating naïve and memory T cells subsets of high competitive level swimmers. Blood samples were taken at rest at 4 time-points during the season: before the start of the season (t0-September), after 7weeks of an initial period of gradually increasing training load (t1-November), after 6weeks of an intense training cycle (t2-February) and 48h after the main competition (t3-April) and from eleven non-athlete controls at 2 similar time-points (t2 and t3). CD4, CD8 and gamma-delta (γδ) T cells expressing the naïve (CCR7(+)CD45RA(+)), central-memory (CM-CCR7(+)CD45RA(-)), effector-memory (EM-CCR7(-)CD45RA(-)) and terminal effector (TEMRA-CCR7(-)CD45RA(+)) were quantified by flow cytometry. Statistical analyses were performed using multilevel modeling regression. Both T CD4(+) naïve and CM presented a linear increase in response to the first moment of training exposure, and had an exponential decrease until the end of the training exposure. As for TCD4(+) EM, changes were observed from t2 until the end of the training season with an exponential trend, while TCD4(+) TEMRA increased linearly throughout the season. TCD8(+) naïve increased at t1 and decreased exponentially thereafter. TCD8(+) TEMRA values decreased at t1 and increased exponentially until t3. γδT-EM had an increase at t1 and an exponential decrease afterwards. In contrast, γδT-TEMRA decreased at t1 and exponentially increased during the remaining 20weeks of training. An increase in TEMRA and EM T cells alongside a decrease in naïve T cells could leave athletes more susceptible to illness in response to variation in training stimulus during the season.