SKB and several other waste management organizations have established the international SKB Task Force on Modeling of Groundwater Flow and Transport of Solutes (TF GWFTS) to support and interpret field experiments. Objectives of the task force are to develop, test and improve tools for conceptual understanding and simulating groundwater flow and transport of solutes in fractured rocks. Work is organized in collaborative modeling tasks. Task 9 focuses on realistic modeling of coupled matrix diffusion and sorption in heterogeneous crystalline rock matrix at depth, e.g. by inverse and predictive modeling of in-situ transport experiments. Posiva’s REPRO (rock matrix REtention PROperties) experimental campaign has been performed at the ONKALO rock characterization facility in Finland. The two REPRO experiments considered were the Water Phase Diffusion Experiment (WPDE), addressing matrix diffusion in gneiss around a single borehole interval (modeled in Task 9A), and the Through Diffusion Experiment, which is performed between sections of three boreholes and addressed by modeling in Task 9C. The Long-Term Diffusion and Sorption Experiment (LTDE-SD) was an in-situ radionuclide tracer test performed at the Swedish Äspö Hard Rock Laboratory at a depth of about 410 m below sea level. The experimental results indicated a possible deeper penetration of sorbing tracers into the rock matrix than expected. The shape of these tracer penetration profiles was difficult to reproduce. This experiment was modeled and interpreted in Task 9B. Task 9D is addressing the possible benefits of detailed models of the in-situ experiments in safety assessment calculations. The task is performed by upscaling of the WPDE models to conditions applicable for nuclear waste repositories. As Task 9 is now in a finalization process, a number of lessons learned from the 4 sub-tasks have been identified. These include: • field tracer experiments can provide surprises even when well designed and executed, • interaction between the experimentalists and modelers is important and mutually beneficial when investigating anomalous results, • differences in conceptual models have the greatest impact on model outcomes, • it is not trivial to go from modeling of field experiments to safety assessment modeling without making substantial simplifications.