Multiphysics problems in complex hydrology systems including the interaction of surface and subsurface waters have been already recognized as typical examples where separate analysis does not present realistic and acceptable solutions [1]. Very important examples of such processes are flow and transport in karst aquifers. Conceptual karst model usually presents 3-D porous matrix, 2-D overland flow and 1-D conduits [2]. Particular challenge remains in physically correct formulations of coupling between surface and subsurface components. Especially, tracer transport is sensitive due to unknown heterogeneity of system, inaccurate flow solutions where small velocity errors can ruin tracer results as well as verification and validation of computational karst models. Particular contribution is based on verification of karst computational methods using 3-D controlled laboratory experiments in new Hydrolab in University of Split [2]. 3-D physical model, filled by quartz sand and perforated pipes (conduits), is build using extensive measurement resources which is still not possible in realistic systems. Different computational models are verified by salt tracer tests under different flow conditions: rain, wet or drought conditions, different conduit setups or different piezometric relationship between matrix and conduits. Results present that transport coupling using only advective exchange fluxes enable satisfactory matching between experiments and adequate karst transport models. It can be serve as solid platform for realistic catchment hydrology systems.