In this paper, after describing a global methodology to simulate stochastically 3D karstic networks, we present different strategies to integrate speleogenetic information, geological knowledge and field data that may be available, such as inception horizons, (paleo)water-table, faults, orientation and density of fractures, inlets, outlets, tracer test results. Connectivity characteristics of these networks can also be taken into account. The promising results open new avenues for the simulation of cave systems in 3D.