Abstract: Based on correlations of summer temperature over Central Asia with spring sea surface temperature, spring soil temperature, and spring large-scale climate indexes for the period of 1979-2016, the study identifies four spring predictors including sea surface temperature over the southeastern Indian Ocean, soil temperature over northwestern Africa, the Atlantic Multidecadal Oscillation (AMO) and the Eastern Asia/Western Russia (EA/WR) pattern to establish a prediction model for summer temperature over Central Asia. Positive anomalies of spring sea surface temperature over the southeastern Indian Ocean, spring soil temperature over northwest Africa and spring AMO and negative anomalies of spring EA/WR are corresponding to positive summer geopotential height anomalies over Central Asia, which are favorable for high summer temperature. The correlation coefficient between the detrended (original) time series of summer temperature averaged over Central Asia produced by leave-one-out cross-validation and the observation is 0.65 (0.74), indicating that the model has a good performance in seasonal prediction of summer temperature over Central Asia. Results of the present study are expected to help improve the seasonal prediction skill of summer temperature over Central Asia.