Spatial and temporal temperature and moisture patterns across the Tibetan Plateau are very complex. The onset and magnitude of the Holocene climate optimum in the Asian monsoon realm, in particular, is a subject of considerable debate as this time period is often used as an analogue for recent global warming. In the light of contradictory inferences regarding past climate and environmental change on the Tibetan Plateau, I have attempted to explain mismatches in the timing and magnitude of change. Therefore, I analysed the temporal variation of fossil pollen and diatom spectra and the geochemical record from palaeo-ecological records covering different time scales (late Quaternary and the last 200 years) from two core regions in the NE and SE Tibetan Plateau. For interpretation purposes I combined my data with other available palaeo-ecological data to set up corresponding aquatic and terrestrial proxy data sets of two lake pairs and two sets of sites. I focused on the direct comparison of proxies representing lacustrine response to climate signals (e.g., diatoms, ostracods, geochemical record) and proxies representing changes in the terrestrial environment (i.e., terrestrial pollen), in order to asses whether the lake and its catchments respond at similar times and magnitudes to environmental changes. Therefore, I introduced the established numerical technique procrustes rotation as a new approach in palaeoecology to quantitatively compare raw data of any two sedimentary records of interest in order to assess their degree of concordance. Focusing on the late Quaternary, sediment cores from two lakes (Kuhai Lake 35.3°N; 99.2°E; 4150 m asl; and Koucha Lake 34.0°N; 97.2°E; 4540 m asl) on the semi-arid northeastern Tibetan Plateau were analysed to identify post-glacial vegetation and environmental changes, and to investigate the responses of lake ecosystems to such changes. Based on the pollen record, five major vegetation and climate changes could be identified: (1) A shift from alpine desert to alpine steppe indicates a change from cold, dry conditions to warmer and more moist conditions at 14.8 cal. ka BP, (2) alpine steppe with tundra elements points to conditions of higher effective moisture and a stepwise warming climate at 13.6 cal. ka BP, (3) the appearance of high-alpine meadow vegetation indicates a further change towards increased moisture, but with colder temperatures, at 7.0 cal. ka BP, (4) the reoccurrence of alpine steppe with desert elements suggests a return to a significantly colder and drier phase at 6.3 cal. ka BP, and (5) the establishment of alpine steppe-meadow vegetation indicates a change back to relatively moist conditions at 2.2 cal. ka BP. To place the reconstructed climate inferences from the NE Tibetan Plateau into the context of Holocene moisture evolution across the Tibetan Plateau, I applied a five-scale moisture index and average link clustering to all available continuous pollen and non-pollen palaeoclimate records from the Tibetan Plateau, in an attempt to detect coherent regional and temporal patterns of moisture evolution on the Plateau. However, no common temporal or spatial pattern of moisture evolution during the Holocene could be detected, which can be assigned to the complex responses of different proxies to environmental changes in an already very heterogeneous mountain landscape, where minor differences in elevation can result in marked variations in microenvironments. Focusing on the past 200 years, I analysed the sedimentary records (LC6 Lake 29.5°N, 94.3°E, 4132 m asl; and Wuxu Lake 29.9°N, 101.1°E, 3705 m asl) from the southeastern Tibetan Plateau. I found that despite presumed significant temperature increases over that period, pollen and diatom records from the SE Tibetan Plateau reveal only very subtle changes throughout their profiles. The compositional species turnover investigated over the last 200 years appears relatively low in comparison to the species reorganisations during the Holocene. The results indicate that climatically induced ecological thresholds are not yet crossed, but that human activity has an increasing influence, particularly on the terrestrial ecosystem. Forest clearances and reforestation have not caused forest decline in our study area, but a conversion of natural forests to semi-natural secondary forests. The results from the numerical proxy comparison of the two sets of two pairs of Tibetan lakes indicate that the use of different proxies and the work with palaeo-ecological records from different lake types can cause deviant stories of inferred change. Irrespective of the timescale (Holocene or last 200 years) or region (SE or NE Tibetan Plateau) analysed, the agreement in terms of the direction, timing, and magnitude of change between the corresponding terrestrial data sets is generally better than the match between the corresponding lacustrine data sets, suggesting that lacustrine proxies may partly be influenced by in-lake or local catchment processes whereas the terrestrial proxy reflects a more regional climatic signal. The current disaccord on coherent temporal and spatial climate patterns on the Tibetan Plateau can partly be ascribed to the complexity of proxy response and lake systems on the Tibetan Plateau. Therefore, a multi-proxy, multi-site approach is important in order to gain a reliable climate interpretation for the complex mountain landscape of the Tibetan Plateau.