The results from stratigraphy, radiocarbon dating and pollen analysis of three salt-lake deposits at Sambhar, Lunkaransar and Didwana in western Rajasthan, and one freshwater lake deposit at Pushkar in the Aravalli Hills, are described in conjunction with pollen analysis of some archaeological soil samples from the Indus Valley site at Kalibangan in northern Rajasthan. The salt-lake deposits studied are stratigraphically divisible into (a) pre-lacustrine, (b) lacustrine and (c) post-lacustrine sections. The pre-lacustrine section is characterized by a thick bed of aeolian sand underlying lacustrine sediments, while the lacustrine and post-lacustrine sections are broadly circumscribed by laminated clay and nonlaminated silt respectively. The pollen record from the four lake profiles studied is divided into local pollen zones. Four regional pollen assemblage zones are delineated for the area west of the Aravalli Range in Rajasthan. The environmental history deduced from the pollen record is divisible into phases I-V, of which phases II-V follow the regional pollen assemblage zones. Phase I is stratigraphically determined, and is representative of severe arid environments under which the sand dunes, now stabilized, are suggested to have been active. The plant microfossils first appear in phase II with the deposition of lacustrine sediments dated to around 10000 B.P. The vegetation comprises an openland steppe which is rich in grasses, Artemisia and sedges and poor in halophytes. Artemisia, Typha angustata, Mimosa rubicaulis and Oldenlandia, which now grow under areas of comparatively higher average annual rainfall (above 50 cm), appear to have flourished in the semi-arid belt, while the first two plant taxa had encroached even as far as the arid belt, both suggesting that a general westward shift of the rainfall belts had taken place. Vegetation destruction through burning by man is suggested to have started together with the first occurrence of Cerealia-type pollen at about 7500 B.C. and continued thereafter throughout phases III and IV. The increase in swamp vegetation and the intensification of vegetation cover inland together with the maxima of all mesophytic elements in phase IV (ca. 3000 B.C. to ca. 1000 B.C.) indicate an increase in the rainfall, apart from a short relatively drier time about 1800-1500 B.C. at Sambhar which correlates with the decline of the Indus culture in northwest India. Phase IV is immediately followed by aridity for which there is stratigraphic evidence that the salt lakes started drying. At Pushkar, there is evidence that the vegetation showed a marked change in the Aravallis. The onset of this aridity is suggested to have been widespread. The climate did not ameliorate until about phase V (? early centuries A.D. to present) at which time the Rangmahal culture perhaps flourished in Rajasthan, the remains of which imply good water supply. In conclusion it is suggested that the Rajasthan desert is primarily natural, its history punctuated by at least one more vegetated, humid period during the Holocene, the climatic control of which as indicated by the vegetation history is consistent with climatic events elsewhere in the world.