The response of soil wax derived n-alkane hydrogen isotope values (δDwax, the abundance weighted average δD values of the C29 and C31n-alkanes) and isotopic lapse rates (the average change in δDwax with elevation) to factors other than elevation change should be assessed and taken into account for quantitative elevation reconstructions. To shed more light on how plant wax derived n-alkanes may be used for this purpose, soil and Gramineae samples analyzed for δDwax values. Those were derived from the foreland basin, the mountain vegetation zone and glacier catchment areas along the Beida River (a tributary in the Heihe River Basin; up to the headwaters of the Qiyi glacier), and the Xiying River catchment (up to the headwaters and glaciated areas of the Lenglong Range), located on the northeastern Tibetan Plateau. For comparison, this study analyzed surface soils from the Maxian Mountains for δDwax. As a whole, the trend in δDwax in the study regions seems to primarily reflect δDrw (δD values of river water) and/or δDp (δD values of precipitation). These data show good inverse linear relationships between Gramineae and soil δDwax values and elevation (altitude effect) in the mountain vegetation zone of the Xiying River catchment (2200–3500m) and Maxian Mountains (2800–3700m), where the elevation effect becomes the main control factor even in the dry Qilian Mountains. However, at the low elevation sites, soil and Gramineae δDwax values from the Xiying River (1750–2200m) and Maxian Mountains (2060–2800m) show the “inverse altitude effect”. The recycled fraction and trajectory of water vapor may result in these positive δDwax elevation gradients in dry continental settings. The “inverse altitude effect” has also been observed in periglacial regions below the Qiyi glacier and from the Lenglong Range. We interpret this effect as evidence of stronger fractionation processes occurring in the freeze–thaw mass exchange between the solid and liquid phases at high elevations. Thus, the “inverse altitude effect” can be attributed to the decrease of precipitation amount in the upper valley and to marked fractionation in soil water and the residual or melting snow in periglacial regions, which may complicate the isotope thermometry. When interpreting changes of long term δD and δ18O records for determining elevation change, one must consider all possible influences on the lapse rate of δDwax values with elevation, including the patterns of atmospheric circulation in the past, topography, recycling of moisture in dry continental settings, and strong fractionation processes effecting soil and glacier melt water in the periglacial regions.