Question of the study On the basis of polysomnographic laboratory and field studies, the DLR Institute of Aerospace Medicine has developed a concept to protect against adverse effects of nocturnal aircraft noise at Airport Leipzig/Halle, which will be extended to a freight hub. We investigated whether or not the expected high traffic densities during the night will relevantly interfere with sleep macrostructure, if the criteria suggested by DLR are met. Methods Models were based on data sampled in a polysomnographic field study on 64 residents living in the vicinity of Airport Cologne/Bonn. Markov processes were used to model a time in bed (TIB) of 8 h. Only two states were differentiated: wake and sleep, the latter consisting of S1-S4 and REM. Transition probabilities were estimated with logistic regression including current state (wake/sleep), duration of current state, elapsed sleep time, and maximum sound pressure level (SPL) of the aircraft noise event (ANE) as covariates. Three traffic scenarios were simulated: (i) no noise, (ii) aircraft noise at the beginning of the night (model A), and (iii) aircraft noise at the end of the night (model E). Results On the basis of 8 hours TIB and compared to the scenario without aircraft noise (81.1 min), time spent awake increased by 3.8 min (+4.7 %) in model A and by 5.9 min (+7.3 %) in model E. However, aircraft noise at the end of the night caused more pronounced changes in sleep structure than aircraft noise at the beginning of the night. Therefore, we propose to levy a malus of 1.4 dB on aircraft noise events occurring in the second half of the night. Conclusion In the context of noise effects on sleep, Markov processes allow for a flexible modelling of dependent events and variable traffic scenarios. If the criteria of the DLR protection concept are met, the models predict only minor noise-induced changes in time spent awake and in the number of awakenings recalled in the morning. [ABSTRACT FROM AUTHOR]