Abstract: The scale feature of the Xinjiang climate change associated with the recent 20- year warm-wet climate is studied by wavelet decomposition. One of regional characteristics in climatology is the collocation of temperature and rainfall. There are four types of collocations i.e. warm-wet, cool-dry, warm-dry and cool-wet climates. As well known, a rainy climate does not represent a wet climate sometimes due to surface evaporation change. For example, if the increase rate excess the precipitation the climate would become drier. For avoiding miss-understanding, the words warm-rainy or cool-rainy are used in following content of this paper. Xinjiang is a Uygur autonomous region in northwestern China with a very dry continent climate. The water resource is one of the most important factors for regional economic development. In contrary to the severe drought occurred in north China during the past 30-year, the rainfall in Xinjiang has apparently increased since mid-1980s accompanying a warm climate under the background of global warming. Some of scientific researches conjecture that a climate transition from warm-dry to warm-rainy has occurred in mid-1980s. This change has led to more rainfall and warm climate there, mountain glacial shrinkage, runoff augment, lake expansion, frequent floods and so on. The surface vegetation has been ameliorated significantly according to remote sensing from satellite. The rainfall wavelet power spectrum shows 2-4-years /and 6-8-years periods at inter-annual variation belt, and about 16-year period at inter-decadal scale. The different spectra of the temperature and rainfall sequences imply a complex collocation with their evolutions. In addition to continuous wavelet analysis the temperature and rainfall sequences are also decomposed with orthogonal wavelets Daub4 for reducing the edge-effect. It shows a significantly negative correlation between them at high frequency belt and a weak correlation in low frequencies within inter-annual variation belt, resulting from their non-stationary properties. A further quantitative investigation shows that the major precipitation energy concentrates on inter-annual variation and the part beyond 50-year components, as well, while the trend is the dominant component in temperature decomposition. The temperature and rainfall components exhibit a significant correlation, i.e. a warm-rainy/cool-dry climate collocation for the components. A further analysis has revealed that the appearance of the Xinjiang warm-rainy climate is contributed by the positive phase of 50-60-year component in precipitation and the warm trend during the recent 20-year. The persistence of the warm-rainy climate in Xinjiang seems expected in near future if global warming keeps on going. Finally, uncertainties for such a simple estimate to future climate of Xinjiang is discussed with the consideration of sea-ice interaction over west Europe and the Atlantic.