This research investigates the dispersion of colloids through fracture systems by exploring experimentally and numerically the transport and dispersion of 1.0-, 0.11-, and 0.043-mum diameter fluorescent carboxylate-modified microspheres and chloride at various flow rates through variable-length, synthetic Plexiglas fractures (flow cells). A dimensionless number describing each experiment is varied by changing the colloid size, flow rate, and fracture length. Surface characteristics of the microspheres and Plexiglas favor repulsive interactions, thereby minimizing the chance of colloid filtration and remobilization. Full recovery of the colloids is typically observed, thereby supporting the assumption of negligible colloid filtration. In comparison to chloride transport, there is increased tailing for colloid plumes traveling through the flow cell. This increased tailing is attributed to Taylor dispersion phenomena (dispersion due to an advection gradient). In the synthetic fractures investigated here, colloid dispersion due to the velocity gradient is evident, but fully developed Taylor conditions are not realized. A particle-tracking algorithm is run inversely to estimate the effective dispersion rate for the colloid plume in each experiment as a function of the experimental parameters (flow rate, fracture length, and colloid size). Results suggest that the log of the effective dispersion rate of the colloid plume increases linearly with the log of the dimensionless number comprising experimental parameters.