Prefrontal cortex (PFC)-related functions are particularly sensitive to sleep loss. However, their repeated examination is intricate because of methodological constraints such as practice effects and loss of novelty. We investigated to what extent the circadian timing system and the sleep homeostat influence PFC-related performance in differently difficult versions of a single task. Parallel versions of a planning task combined with a control group investigation were used to control for practice effects. Thirteen healthy volunteers (five women and eight men, range 57-74 years) completed a 40-h sleep deprivation (SD) and a 40-h multiple nap protocol (NAP) under constant routine conditions. Each participant performed 11 easy and 11 difficult task versions under either SD or NAP conditions. The cognitive and motor components of performance could be distinguished and analysed separately. Only by thoroughly controlling for superimposed secondary factors such as practice or sequence effects, could a significant influence of circadian timing and sleep pressure be clearly detected in planning performance in the more difficult, but not easier maze tasks. These results indicate that sleep loss-related decrements in planning performance depend on difficulty level, and that apparently insensitive tasks can turn out to be sensitive to sleep loss and circadian variation.