The U.S. Geological Survey (USGS), in cooperation with the Tennessee Department of Environment and Conservation, Division of Superfund, collected discharge, rainfall, continuous water-quality (temperature, dissolved oxygen, specific conductance, and pH), and volatile organic compound (VOC) data from three karst springs in Middle Tennessee from February 2000 to May 2001. Continuous monitoring data indicated that each spring responds differently to storms. Water quality and discharge at Wilson Spring, which is located in the Central Basin karst region of Tennessee, changed rapidly after rainfall. Water quality and discharge also varied at Cascade Spring; however, changes did not occur as frequently or as quickly as changes at Wilson Spring. Water quality and discharge at Big Spring at Rutledge Falls changed little in response to storms. Cascade Spring and Big Spring at Rutledge Falls are located in similar hydrogeologic settings on the escarpment of the Highland Rim. Nonisokinetic dip-sampling methods were used to collect VOC samples from the springs during base-flow conditions. During selected storms, automatic samplers were used to collect water samples at Cascade Spring and Wilson Spring. Water samples were collected as frequently as every 15 minutes at the beginning of a storm, and sampling intervals were gradually increased following a storm. VOC samples were analyzed using a portable gas chromatograph (GC). VOC samples were collected from Wilson, Cascade, and Big Springs during 600, 199, and 55 sampling times, respectively, from February 2000 to May 2001. Chloroform concentrations detected at Wilson Spring ranged from 0.073 to 34 mg/L (milligrams per liter). Chloroform concentrations changed during most storms; the greatest change detected was during the first storm in fall 2000, when chloroform concentrations increased from about 0.5 to about 34 mg/L. Concentrations of cis-1,2-dichloroethylene (cis-1,2-DCE) detected at Cascade Spring ranged from 0.30 to 1.8 ?g/L (micrograms per liter) and gradually decreased between November 2000 and May 2001. In addition to the gradual decrease in cis-1,2-DCE concentrations, some additional decreases were detected during storms. VOC samples collected at weekly intervals from Big Spring indicated a gradual decrease in trichloroethylene (TCE) concentrations from approximately 9 to 6 ?g/L between November 2000 and May 2001. Significant changes in TCE concentrations were not detected during individual storms at Big Spring. Quality-control samples included trip blanks, equipment blanks, replicates, and field-matrix spike samples. VOC concentrations measured using the portable GC were similar to concentrations in replicate samples analyzed by the USGS National Water Quality Laboratory (NWQL) with the exception of chloroform and TCE concentrations. Chloroform and TCE concentrations detected by the portable GC were consistently lower (median percent differences of ?19.2 and ?17.4, respectively) than NWQL results. High correlations, however, were observed between concentrations detected by the portable GC and concentrations detected by the NWQL (Pearson?s r > 0.96). VOC concentrations in automatically collected samples were similar to concentrations in replicates collected using dip-sampling methods. More than 80 percent of the VOC concentrations measured in automatically collected samples were within 12 percent of concentrations in dip samples.