1] A tracer experiment, using a nonreactive tracer, was conducted as part of an investigation of the potential for chemical and pathogen migration to public supply wells that draw groundwater from the highly transmissive karst limestone of the Biscayne aquifer in southeastern Florida. The tracer was injected into the formation over approximately 1 h, and its recovery was monitored at a pumping well approximately 100 m from the injection well. The first detection of the tracer occurred after approximately 5 h, and the peak concentration occurred at about 8 h after the injection. The tracer was still detected in the production well more than 6 days after injection, and only 42% of the tracer mass was recovered. It is hypothesized that a combination of chemical diffusion and slow advection resulted in significant retention of the tracer in the formation, despite the high transmissivity of the karst limestone. The tail of the breakthrough curve exhibited a straight-line behavior with a slope of À2 on a log-log plot of concentration versus time. The À2 slope is hypothesized to be a function of slow advection, where the velocities of flow paths are hypothesized to range over several orders of magnitude. The flow paths having the slowest velocities result in a response similar to chemical diffusion. Chemical diffusion, due to chemical gradients, is still ongoing during the declining limb of the breakthrough curve, but this process is dwarfed by the magnitude of the mass flux by slow advection.