Using long-term daily streamflow records from the headwaters of the Tarim River Basin in northwestern China, the mean seasonal streamflow and extremes (1-day maximum and minimum flows) are analysed. The aim is to assess possible changes in trends, return periods and teleconnections to selected climate indices such as the North Atlantic Oscillation (NAO), the Arctic Oscillation (AO), the multivariate El Nino–Southern Oscillation index (MEI), the Tibetan Plateau High (XZH), the Northern Hemisphere polar vortex area (VPA), the Asian zonal index (AZI) and the Siberian High intensity index (SHI). The Mann–Kendall test is used to investigate instances where significant positive trends are observed in all hydrological indices, especially in winter. From 1990 onwards, a significant wavelet power spectrum of annual and seasonal flow was observed at 1–5 year periods in the Aksu River. Correlation analysis and wavelet coherence analysis were employed between selected climate indices with respect to seasonal streamflow means and extremes. Maximum changes in correlation strengths were observed in hydrological indices with VPA. The winter AO and NAO also showed significant influence on streamflow means and extremes. In addition, step change analysis was carried out using the Cumulative Sum Control Chart (CUSUM) test, the results suggesting that step change points were observed towards the end of the 1980s and in the early 1990s. The teleconnection between climate indices and streamflow and extremes can be utilized in a model to predict the probability of hydrological indices in the Tarim River.