Compound specific hydrogen isotope ratios (δD) of long chain sedimentary n-alkanes, which mostly originate from the leaf waxes of higher terrestrial plants, are increasingly employed as paleoclimate proxies. While soil water is the ultimate hydrogen source for these lipids and the isotopic fractionation during biosynthesis of lipids is thought to remain constant, environmental parameters and plant physiological processes can alter the apparent hydrogen isotopic fractionation between leaf-wax lipids and a plant’s source water. However, the magnitude and timing of these effects and their influence on the isotopic composition of lipids from higher terrestrial plants are still not well understood. Therefore we investigated the seasonal variability of leaf-wax n-alkane δD values for two different temperate deciduous forest ecosystems that are dominated by two different tree species, Beech (Fagus sylvatica) and Maple (Acerpseudoplatanus). We found significant seasonal variations for both tree species in n-alkane δD values of up to 40‰ on timescales as short as one week. Also, the isotopic difference between different n-alkanes from the same plant species did vary significantly and reached up to 50‰ at the same time when overall n-alkane concentrations were lowest. Since δD values of soil water at 5 and 10cm depth, which we assume represent the δD value of the major water source for the investigated beech trees, were enriched in autumn compared to the spring by 30‰, whereas n-alkane δD values increased only by 10‰, we observed variations in the apparent fractionation between beech leaf derived n-alkanes and soil water of up to 20‰ on a seasonal scale. This observed change in the apparent fractionation was likely caused by differences in leaf water isotopic enrichment. Based on mechanistic leaf water models we conclude that changes in the isotopic difference between water vapor and soil water were the most likely reason for the observed changes in the apparent fractionation between n-alkanes and soil water. The large variability of n-alkane concentrations and δD values over time implies a continuous de novo synthesis of these compounds over the growing season with turnover times possibly as short as weeks. The signal to reach the soil therefore represents an integrated record of the last weeks before leaf senescence. This holds true also for the sedimentary record of small catchment lakes in humid, temperate climates, where wind transport of leaf-wax lipids is negligible compared to transfer through soil and the massive input of leaves directly into the lake in autumn.