AbstractEmpirical statistical downscaling has been widely used to produce finer‐resolution climate data. This approach, in general, is derived from an implicit stationarity assumption. This paper aims at proving a statistical method that is fully applicable of generating daily precipitation in non‐stationary conditions using historical station data. Daily records at five Oklahoma stations were split into calibration and validation periods. Linear relationships between transition probabilities of wet‐following‐wet (Pw/w) and wet‐following‐dry (Pw/d) days and mean monthly precipitation were established by connecting the two endpoints (one for the 30 driest months and another for the 30 wettest months of the calibration period) for each calendar month, and were then used to interpolate Pw/w and Pw/d for the validation period. The mean and standard deviation of daily precipitation were estimated using the mean and standard deviation of monthly precipitation of the validation periods as well as the interpolated Pw/w and Pw/d. The adjusted parameters were used to generate daily series using a weather generator. Statistics of the disaggregated daily precipitation amounts and frequency, as well as dry/wet spell sequence, agreed with those of the observed values of the validation periods reasonably well. The disaggregation method preserved statistics of monthly precipitation amounts extremely well, demonstrating the validity of the method for temporal disaggregation of non‐stationary climate. The accuracy of the presented method is within the weather generator's expected performance. Overall, a straight line connecting the two endpoints, implicitly incorporating non‐stationarity of climate states, is adequate for interpolating Pw/w and Pw/d to any climatic conditions within the entire range. However, a linear regression including data points in between would generally improve the Pw/w and Pw/d interpolations slightly if long records (e.g. > 60 years) are available for estimating the intermediate points. Copyright © 2012 Royal Meteorological Society