Inspired by recently published chloride balance chronologies of closed-basin lakes in the Great Basin, researchers of the early 1960s attempted to estimate residence times of chloride in these lakes using 36Cl. Unfortunately, the analytical methods of the period were not capable of measuring the 36Cl levels found in these waters. About 20 years after the early research, advances in accelerator mass spectrometry permitted 36Cl measurement at the required sensitivity. In this study we follow up on those pioneering efforts by remeasuring and reevaluating the 36Cl content at several of the previously studied sites, focusing on Mono Lake in eastern California. Our data show that in general the streams in the region have 36Cl/Cl ratios similar to those expected in present-day atmospheric fallout, but that the terminal lakes into which the streams flow have much lower ratios. These lower ratios could result from either a very long (>1 million years) residence time of the chloride in the basin sinks or from subsurface influx of low-36Cl chloride. In the case of Mono Lake, a mass balance model based on the 36Cl data and on independent estimates of chloride fluxes and reservoirs indicates major subsurface chloride input, presumably from volcanic sources, and an accumulation time in the range of 100--450 kyr. The upper bound of this range is similar to the timing of a shift from long-term humid to arid climate in the region and may indicate that hydrological closure of the basin was triggered by this event. ¿ American Geophysical Union 1995