Hydrologic drought can be studied in different ways. One of the common methods is the use of low flow indexes. In this study for the purpose of determine of the identity of aspects of low flow, the homogenous zones, the extraction of regional regression models, and finally, the study of low flow trends of Karkheh river basin, applied the data of 13 hydrometric stations during the statistical period of 1960-2000. After qualitative and quantitative controlling and the retrieval of missing data used flow duration curves for indexes such as Q75, Q90 and Q95. Also frequency analysis of 10-day, and 20-day low flows was carried out for indexes such as Q10, 5, Q10, 50, Q10, 100, Q30, 5, Q30, 50, Q30, 100 and some more indexes. Afterwards, some other factors like physiographic, climatic, geologic and vegetation cover were applied as influential parameters in the regional analysis. These factors were used in cluster analysis and stepwise regression estimations. Final step was the trend analysis of times series of the indexes. Results indicate that among the indexes being calculated, Q10, 100 and Q95 had the minimum quantities, whose amounts reduced as the return periods increased. On the hand, the time spans of 1998-1999, 1999-2000, 2000-2001 have undergone severe and long droughts in most of the stations. The review of the spatial distribution of indexes show better conditions of the south-eastern parts of the study area compared to the northern and southern sections in terms of dryness. Results of cluster analysis divided the area into two distinct homogenous units (in 0. 01 significant level). In the area No. 1, the elevation factor, in the area No. 2, the drainage density, and in general, the factors such as the mean of area and drainage density have the highest effects. The Spearman statistic, and Mann-Kendall findings also indicate that the low flow in upper basin have negative trend during the statistical period.