The literature on probability techniques applicable to problems in hydrology is abundant but scattered through scientific journals of both hydrology and statistics. Important administrative and judicial decisions presently face water-compact commissions, courts, and water-planning committees. These and other groups might find useful, a brief and simplified discussion of how statistical techniques can aid in analysing problems of water supply. The interest expressed in this subject by various parties to the litigation concerning the Colorado River prompts this publication of material, which was presented in August 1958 before the Special Master of the Supreme Court hearing the proceedings of Arizona v. California et al. The examples presented here are the same as those used in testimony before the Special Master, but there are included the basic computations, which were too detailed to present in the actual testimony. The specific example, which was analyzed in that testimony, was a 61-year series of annual discharge values of the Colorado River at Lees Ferry, 1896 to 1956, inclusive. However, the methodology presented herein is generally applicable to many other streamflow records; and the specific data discussed should be viewed as exemplifying the types of information, which can be obtained from any streamflow record.