The suitability of passive capillary samplers (PCAPS) for collection of representative colloid samples under partially saturated conditions was evaluated by investigating the transport of negatively and positively charged colloids in fiberglass wicks. A synthetic pore water solution was used to suspend silica microspheres (330 nm in diameter) and ferrihydrite (172 nm in diameter) for transport experiments on fiberglass wicks. Breakthrough curves were collected for three unsaturated flow rates with silica microspheres and one unsaturated flow rate with ferrihydrite colloids. A moisture characteristic curve, relating tensiometer measurements of matric potential to moisture content, was developed for the fiberglass wick. Results indicate that retention of the silica and the ferrihydrite on the wick occurred; that is, the wicks did not facilitate quantitative sampling of the colloids. For silica microspheres, 90% of the colloids were transmitted through the wicks. For ferrihydrite, 80 to 90% of the colloids were transmitted. The mechanisms responsible for the retention of the colloids on the fiberglass wicks appeared to be physicochemical attachment and not thin-film, triple-phase entrapment, or mechanical straining. Visualization of pathways by iron staining indicates that flow is preferential at the center of twisted bundles of filaments. Although axial preferential flow in PCAPS may enhance their hydraulic suitability for sampling mobile colloids, we conclude that without specific preparation to reduce attachment or retention, fiberglass wicks should only be used for qualitative sampling of pore water colloids.