Low-frequency (interannual or longer period) climatic variability is of interest because of its significance for the understanding and prediction of protracted climtic anomalies. Closed basin lakes are sensitive to long-term climatic fluctuations and integrate out high-frequency variability. It is thus natural to examine the records of such lakes to better understand long-term climate dynamics. Here we use singular spectral analysis and multitaper spectral analysis to analyze the time series of Great Salt Lake (GSL) monthly volume change from 1848 to 1992 and monthly precipitation, temperature, and streamflow for nearby stations with 74 or more years of data. This analysis reveals high fractional variance in 15--18, 10--12, 3--7 and 2-year frequency bands, which seems to be consistent across time series. The putative decadal and interdecadal signals appear to be related to large-scale climate signals. The interannual signals are consistent with El Ni¿o Southern Oscillation and quasi-biennial variability. Prospects for improved prediction of the GSL volume and of protracted wet/dry periods in the western United States are discussed. ¿ American Geophysical Union 1995.