Environmental tracers (noble gases, tritium, industrial gases, stable isotopes, and radio-carbon) and hydrogeology were interpreted to determine groundwater transit-time distribution and calculate mean transit time (MTT) with lumped parameter modeling at 19 large springs distributed throughout the Upper Colorado River Basin (UCRB), USA. The predictive value of the MTT to evaluate the pattern and timing of groundwater response to hydraulic stress (i.e., vulnerability) is examined by a statistical analysis of MTT, historical spring discharge records, and the Palmer Hydrological Drought Index. MTTs of the springs range from 10 to 15,000 years and 90 % of the cumulative discharge-weighted travel-time distribution falls within the range of 2−10,000 years. Historical variability in discharge was assessed as the ratio of 10–90 % flow-exceedance (R _10/90%) and ranged from 2.8 to 1.1 for select springs with available discharge data. The lag-time (i.e., delay in discharge response to drought conditions) was determined by cross-correlation analysis and ranged from 0.5 to 6 years for the same select springs. Springs with shorter MTTs (<80 years) statistically correlate with larger discharge variations and faster responses to drought, indicating MTT can be used for estimating the relative magnitude and timing of groundwater response. Results indicate that groundwater discharge to streams in the UCRB will likely respond on the order of years to climate variation and increasing groundwater withdrawals.