Synthesized monthly natural streamflows were required at 13 sites for use in a streamflow- accounting model to evaluate the effects of various water-allocation schemes on water availability in the Musselshell River Basin in central Montana. Records of monthly streamflow at 14 streamflow-gaging stations were used to synthesize monthly natural flows at tributaries and the 13 synthesis sites. A streamflow-record extension program was used to extend flow records at the 14 gaged sites to a common base period, 1929-89. To synthesize monthly natural flows at 10 sites on the Musselshell River mainstem, synthesized monthly natural flows at all signi- ficant tributary streams were required. Results from a previous study were used to synthesize tributary natural flows. Monthly natural flows at each mainstem site downstream from the first site were synthesized by successively adding monthly natural flows from intervening tributaries to the next upstream mainstem site. Special methods using extended-record flows from gaged tributaries were used to synthesize monthly natural flows at three tributary sites. Synthesized mean annual natural flows were found to be greater than mean annual extended-record flows at three selected comparison sites on the Musselshell River. The differences between mean natural and extended-record flows (depletions) at Harlowton and Musselshell were considered to be reasonable given the amount of irrigated acreage upstream from the two sites. The differences at Mosby, the site farthest downstream, was less than at Musselshell, the next upstream site, indicating that the methods of synthesis had error. The synthesis error generally was attributed to the larger natural variability of tributary flows in the lower portion of the Musselshell River Basin.