Mood, as measured by the profile of mood states questionnaire (POMS), salivary cortisol (F) and testosterone (T) levels, and performance were examined in 17 male soccer players 4 times during a season. Soccer players provided three saliva samples when getting up (resting values, 8 a.m.), before breakfast (11.30 a.m.), and between 4.00 p.m. and 6.00 p.m. The initial measures were performed 1 day following the start of season training (T1). They were then performed before and after a high-intensity training programme (T2 and T3, respectively) and 16 weeks after T3 (T4). Iceberg profiles of POMS were observed during T1, T2 and T3, which coincided with successful performance. Subsequent decreased performance between T3 and T4 coincided with a decrease in vigor and an increase in tension and depression. Indeed, when the normal nycthemeral rhythm for F was observed (i.e. a decrease from morning until evening at all times; T1-T4), there was seemingly a non-statistical elevation of F on the morning of T3, which only became statistically significant at 11.30 a.m. on T3. In spite of a post-high-intensity training programme (T3) increase in catabolism, the soccer players presented iceberg profiles together with a high percentage of winning. Our results could suggest that in team soccer, a decreased T:F ratio does not automatically lead to a decrease in team performance or a state of team overtraining. It appears that combined psychological and physiological changes during high-intensity training are primarily of interest when monitoring training stress in relation to performance.