Abstract: The paper calculated heat source and heat sink over the Tibetan Plateau and its vicinity (QTP) during 1961-2001 by use of the daily ECMWF reanalysis and the “inverse algorithm”, and discussed the climate regimes linked to the thermal sources over the QTP. Results suggest that （1）The region over the QTP with the height more than 3000 m above the sea level acts as a heat source, the maximum is in June with the amount of 214 W/m2 and as a heat sink in October – February with the minimum of –84 W/m2 in December. The heat source lasts for 7 months in the whole atmospheric extent and much stronger than the sink in the wintertime. Therefore the heating effects of the QTP are asymmetric in the seasons. The term of vertical transport of temperature is the factor making greatest contribution to among the three.（2） As shown in the heating vertical profile, the maximum heat source layer occurs dominantly between 500—600 hPa, but with the season-dependent heating strength and depth, and, in contrast, the cold source has its maximum layer and intensity varying as a function of time. The height-varying Q1 is featured mainly by the opposite trend of intensity of the heat to cold source and the whole process can be described as a “cylinder stator” of an engine in operation, with the piston representing source transition that divides the heat and cold source in vertical, as shown in the “steam chests” of the engine. As time goes on, the volume (thickness) is changing constantly for both. As the heat source expands its volume, i.e., the piston goes up, the thickness of the cold source diminishes and v.v. As indicated by the vertical profile, the “piston” reaches its top in July—August when the heat source is the deepest. Conversely, as the “piston” has its lowest position, the cold source covers the greatest depth in October—December, and the troposphere is nearly under the control of the cold source.（3） The horizontal distribution of the heat sources throughout the troposphere (from surface to 100 hPa) is complicated, displaying noticeable regionality, i.e., the heat source changes faster in the western than in the eastern QTP, with the western source considerably stronger in April—August, and intensified quickly enough to show a 200 W/m2 center in May, one month ahead of the eastern source. When July comes the regional heat source begins to weaken towards the south, during which the western source weakens faster, changing to a cold source in October, again one month earlier compared to the eastern counterpart. （4）Since 1979 the seasonal variability of the heat source has shown climate transition signals, as clearly seen in 1990/1991. In summer the variability leads to a see – saw meridionally and for the other seasons there is an anti-phase distribution between the central – NE and SE QTP.