This study investigates the hypothesis that long lasting mental work demands are reflected in after-effects on attention demanding post-test probe tasks. Sixteen subjects were engaged in 2 simulated workdays consisting of mentally demanding tasks. On the difficult day the afternoon consisted of highly demanding information processing tasks; on the easy day the demands during the afternoon were low. The mornings of both days were equally demanding. At the start (pre-test) and the end (post-test) of both days, after 5 min (immediate post-test) as well as after approximately 2 h of recuperation (delayed post-test), a probe task was carried out. A memory-search task (varied mapping) is used as probe task. After-effects were expected in terms of a shift in the ratio between performance and effort on the post-test probe tasks compared with the pre-test probe task. Performance measures were reaction time and errors; effort was indicated by heart rate variability in the 0.07-0. 14 Hz band, referred to as 0.1 Hz component. Dayload manipulation appeared to be effective because of the load-related differences between the urinary excretion rates of adrenaline during both afternoons. Compared with the pre-test, only the number of errors was significantly increased in the post-test, carried out immediately at the end of the difficult afternoon. However, compared with the easy day, in the delayed post-test probe task after the difficult day, subjects tended to shorten reaction times, made significantly more errors, and invested significantly less effort. Moreover, at that time they felt more fatigued and deactivated as indicated by the results of the 'Scale of Experienced Load' (SEB) (fatigue) and 'Groninger Adjective Checklist' (GACL) (mood) questionnaires. The observed shift towards low-effort, more 'risky' task strategies in the post-tests fits well with Hockey's (1993) fatigue after-effects degradation pattern expected after periods of high workload. The repeated measurement of the relation of performance, effort and feelings reveals the dynamics of the pattern of fatigue after-effects over the (recovery) time.