A tracer test in a carbonate aquifer is analyzed using the method of moments and two analytical advection-dispersion models (ADMs) as well as a numerical model. The numerical model is a coupled continuum-pipe flow and transport model that accounts for two different flow components in karstified carbonate aquifers, i.e., rapid and often turbulent conduit flow and Darcian flow in the fissured porous rock. All techniques employed provide reasonable fits to the tracer breakthrough curve (TBC) measured at a spring. The resulting parameter estimates are compared to investigate how each conceptual model of flow and transport processes that forms the basis of the analyses affects the interpretation of the tracer test. Numerical modeling results suggest that the method of moments and the analytical ADMs tend to overestimate the conduit volume because part of the water discharged at the spring is wrongly attributed to the conduit system if flow in the fissured porous rock is ignored. In addition, numerical modeling suggests that mixing of the two flow components accounts for part of the dispersion apparent in the measured TBC, while the remaining part can be attributed to Taylor dispersion. These processes, however, cannot reasonably explain the tail of the TBC. Instead, retention in immobile-fluid regions as included in a nonequilibrium ADM provides a possible explanation.