Dendroclimatology is one of the most important methods for reconstructing past climate change. Owing to the characteristics of precise dating, annual resolution, and comparability with meteorological observational data, tree-ring data have been widely used as important archival material in paleoclimatic research. We selected Siberian spruce (Picea obovata Ledeb) trees with little or no apparent evidence of human or other disturbances for sampling. Tree-ring chronologies at six individual sites were developed for Siberian spruce from the Upper Irtysh River, and then a regional tree-ring chronology (RTC) was established based on all of the detrended data obtained from individual tree cores of the six sites. We analyzed the correlations between the RTC chronology and the meteorological data of the Fuyun weather station. The results showed that precipitation was the main factor limiting the radial growth of spruce trees in this region. The RTC chronology showed the highest correlation with precipitation from the previous July to June of the current year. Based on the results of correlation analysis, we reconstructed annual precipitation patterns (July-June) of the Upper Irtysh River basin from 1722 to 2012. The precipitation reconstruction explained 55.1% of the instrumental precipitation variance during the period of 1963-2012. During the past 291 years, there were nine wet periods and eight dry periods. The wettest period occurred from 1984 to 2008, and the driest period occurred from 1877 to 1891. The period 1829-1876 was the most extended wet period, while the period 1807-1828 was the most prolonged dry period. Power spectrum analysis indicated the existence of some decadal (21.6 and 24.3 year) and interannual (2.1, 2.3 and 3.2 year) cycles. A moving t-test indicated that an abrupt change of precipitation occurred in 1876-1877 and 1983 in this region. The results of spatial correlation analysis indicated that our precipitation reconstruction correlated well (r>0.6) with the July-June precipitation gridded data over a large area of the Altay region, with the highest correlations occurring in the Southern Altay Mountains. When compared with other tree ring based precipitation reconstructions from the surrounding area, our results showed a similar trend in the variation of drought and precipitation.