The objective of this study is to develop a conceptual flow-path model for nuclide migration in fractured crystalline rock at the Kamaishi In-Situ Test Site because improvement of existing models of matrix diffusion |e.g. 1,2|, based on realistic geological data, is necessary for a better understanding of nuclide migration into rock matrix. Data from field observations indicate that fractures at the Kamaishi In-Situ Test Site can be classified into three types; type A with a zone of fracture fillings, type B with a zone of fracture fillings and an altered zone, type C consisting of several fractures with a zone of fracture fillings and an altered zone. Fracture type B was studied in detail by laboratory experiments because type B is predominant in the studied area with more than 60 % of a total of 400 fractures observed in the fracture mapping. Data from laboratory experiments on core, crosscutting a water-bearing fracture and the surrounding rock, indicate that the zone of fracture fillings and the altered zone in the vicinity of the fracture contain flow-paths in which nuclides can migrate and be trapped. The fracture fillings contain more interconnected and permeable flow-paths than the altered and unaltered zones. This implies that migrating nuclides can access flow-paths in the altered zone. The altered zone adjacent to the zone of the fracture fillings contains flow-paths such as microfractures, cracks within quartz, and grain boundaries between altered minerals, through which nuclides will migrate from the fracture fillings into the altered zone and be trapped. The fracture fillings and the specimen of the altered zone have higher sorption capacity than the specimen of the unaltered zone. These data suggest that retention of nuclides can be expected in the vicinity of the fracture. In conclusion, a conceptual flow-path model consisting of a zone of fracture fillings, an altered zone, and an unaltered zone has been developed for a better understanding of nuclide migration in fracture type B.