Strain accumulation and deformation in tectonically active mountain belts are generally well documented over decadal times and on time-scales averaging millions of years. Across timescale data sets, however, are still rare. The intracontinental mountain belts of Central Asia offer a very good possibility to overcome this dilemma because of a relatively long geodetic measurement period, good historical seismic records and a high preservation potential of deformed landscape markers. Observations across instrumental and geologic scales may challenge our concepts of strain localization and accumulation. Shortening across the central Tien Shan, for instance, is homogenous across the entire width of the belt, as seen from geodesy and from averaged Quaternary fault-slip rates showing little variation in deformation rate among slowly-slipping single faults. Preliminary results from basin-wide denudation rates that might be indicative of intermediate-term deformation of the respective ranges show also little variation. Indeed, such a pattern is expected for thick-skinned orogens, like the Tien Shan, where basement-cored uplifts are spatiotemporally disparate and a clear deformation front (as common for thin-skinned provinces) is missing. Contrasting to distributed deformation, the northern rim of the Kyrgyz and Kazakh Tien Shan has experienced a remarkable series of major earthquakes between 1885 and 1938, with a cumulative seismic moment adding to a magnitude of approximately 9. Moderate seismicity still coincides with this well-defined zone; a pattern resembling a deformation front on decadal to perhaps centennial times. Present internal deformation of the Pamir indenter, in turn, seems concentrated along the northern and (north)western margins, virtually irrespective of neighboring thin-skinned and thick-skinned provinces. We will highlight these complexities and discuss how paleoseismic records help to unravel the deformation history over millenial time scales.