The exact manner in which preferential (e.g., much faster than average) flow occurs in the subsurface through small fractures or permeable connected pathways of other kinds is important to many processes but is difficult to determine, because most chemical tracers diffuse quickly enough from small flow channels that they appear to move more uniformly through the rock than they actually do. We show how preferential flow can be assessed by injecting 2 to 5 nm carbon particles (C-Dots) and an inert KBr chemical tracer at different flow rates into a permeable core channel that is surrounded by a less permeable matrix in laboratory apparatus of three different designs. When the KBr tracer has a long enough transit through the system to diffuse into the matrix, but the C-Dot tracer does not, the C-Dot tracer arrives first and the KBr tracer later, and the separation measures the degree of preferential flow. Tracer sequestration in the matrix can be estimated with a Peclet number, and this is useful for experiment design. A model is used to determine the best fitting core and matrix dispersion parameters and refine estimates of the core and matrix porosities. Almost the same parameter values explain all experiments. The methods demonstrated in the laboratory can be applied to field tests. If nanoparticles can be designed that do not stick while flowing through the subsurface, the methods presented here could be used to determine the degree of fracture control in natural environments, and this capability would have very wide ranging value and applicability.