The Central Asian Orogenic Belt (CAOB) is one of the world's largest accretionary orogens, which was active during most of the Paleozoic. In recent years it has again moved into focus of the geological community debating how the acrreted lithospheric elements were geographical arranged and interacting prior and/or during the final amalgamation of Kazakhstania. In principal two families of competing models exist. One possible geodynmaic setting is based on geological evidence that a more or less continuous giant arc connecting Baltica and Siberia in the early Paleozoic was subsequently dissected and buckled. Alternatively an archipelago setting, similar to the present day south west Pacific was proposed.\nThis thesis collates three studies on the paleogeography of the south western part of the CAOB from the early Paleozoic until the latest Paleozoic to earliest Mesozoic. It is shown how fragments of Precambrian to early Paleozoic age are likely to have originated from Gondwana at high southerly paleolatitudes (~500 Ma), which got then accreted during the Ordovician (~460 Ma), before this newly created terrane agglomerate (Kazakhstania) migrated northwards crossing the paleo-equator. During the Devonian and the latest Early Carboniferous (~330 Ma) Kazakhstania occupied a stable position at about ~30\xb0N. At least since this time the area underwent\nseveral stages of counterclockwise rotational movements accompanying the final amalgamation of Eurasia (~320 - ~270 Myr). This overall pattern of\nroughly up to 90\xb0 counterclockwise bending was replaced by internal relative rotational movements in the latest Paleozoic, which continued probably until the early Mesozoic or even the Cenozoic. \nIn Chapter 2 a comparison of declination data acquired by a remagnetization process during folding in the Carboniferous and coeval data from\nBaltica and Siberia lead to a documentation and quantification of rotational movements within the Karatau Mountain Range. Based on this results it\nis very likely that the rotational reorganization started in the Carboniferous and was active until at least the early Mesozoic. Additionally, the data shows that maximal declination deviation increases going from the Karatau towards the Tianshan Mountains (i.e. from North to South). This observation supports models claiming that Ural mountains, Karatau and Tianshan once formed a straight orogen subsequently bent into a orocline. The hinge of this orocline is probably hidden under the sediments of the Caspian basin.\nIn chapter 3 we show that inclination shallowing has affected the red terrigenous sediments of Carboniferous age from the North Tianshan. The\ncorrected inclination values put this part of the Tianshan in a paleolatitude of around 30\xb0N during Carboniferous times. These results contradict previously published paleopositions of the area and suggest a stable latitudinal position between the Devonian and the Carboniferous.\nChapter 4 presents paleomagnetic data from early Paleozoic rocks from within the North Tianshan. They imply a second collisional accretion event\nof individual terranes in the Ordovician.\nTo further constrain the dimensions of these early Paleozoic terranes, chapter 5 presents a compilation of all available paleomagnetic data from the extended study region of southern Kazakhstan and Kyrgyzstan. Apart from a broad coherence of paleolatitudes of all studies at least since the Ordovician and the exclusive occurrence of counterclockwise declination deviations, no areas with the same rotational history can be detected. Also a clear trend caused by oroclinal bending can not be observed. We conclude that first order counterclockwise oroclinal bending, shown in chapter 2, resulted in brittle deformation within the mountain belt and local block rotations.\nIn order to improve our understanding of intra-continental deformation a study combining the monitoring of recent deformation (Global Positioning System, GPS) with a paleomagnetic study of Cenozoic age in the greater vicinity of the Talas-Ferghana fault has been undertaken in chapter 6. The major task was to distinguish between continuous versus brittle deformation.\nAs it turned out the GPS signal indicates rather continuous and consistent counterclockwise rotational movements of the order of ~2\xb0 per Myr. This is in contrast to our paleomagnetic results, where even within fault bounded areas the error intervals of the rotations do always overlap. This indicates that a pure block model seems not appropriate even to explain Cenozoic paleomagnetic\ndata. If this means that also Paleozoic rocks have been affected by complex recent deformation, and that the Paleozoic rotational pattern has been obscured by this, can not be decided based on the present data set. It means, however, that interpreting Paleozoic rotational data from this\narea has to be done with great caution.