Karst aquifers are important resources for drinking water supply and are very vulnerable to contamination. Microbial concentrations at karst springs, in particular, often vary quickly over a short period of time. In this study, the response of microbial water quality and particle-size distribution of two alpine karst springs to rainfall events was investigated to test and validate parameters that can be used as early-warning systems for fecal contamination. At both investigated karst springs, total organic carbon, particle-size distribution (especially small particle fractions), and particle load show a good correlation to the fecal indicator bacteria E. coli and can therefore be used as a real-time indicator of fecal contamination at the investigated springs. In addition to conventional bacterial determination methods, the β-D-glucuronidase activity, which can be measured in near real-time, was used as a novel indicator parameter for fecal contamination. At the event scale, the β-D-glucuronidase (GLUC) activity shows a good correlation to E. coli and can be used as an additional real-time indicator of fecal contamination. For the studied springs, when they show two peaks in turbidity and small particles, these two parameters are suitable for an early warning system because the bacterial contamination occurs during the secondary peak of these parameters. These results highlight the vulnerability of karst aquifers and demonstrate the applicability of advanced measurement techniques in detecting fecal contamination in real-time, which is especially important given the time-consuming nature of conventional bacterial detection methods.