Understanding colloid transport in ground water is essential to assessing the migration of colloid‐size contaminants, the facilitation of dissolved contaminant transport by colloids, in situ bioremediation, and the health risks of pathogen contamination in drinking water wells. Much has been learned through laboratory and field‐scale colloid tracer tests, but progress has been hampered by a lack of consistent tracer testing methodology at different scales and fluid velocities. This paper presents laboratory and field tracer tests in fractured rock that use the same type of colloid tracer over an almost three orders‐of‐magnitude range in scale and fluid velocity. Fluorescently‐dyed carboxylate‐modified latex (CML) microspheres (0.19 to 0.98 μm diameter) were used as tracers in (1) a naturally fractured tuff sample, (2) a large block of naturally fractured granite, (3) a fractured granite field site, and (4) another fractured granite/schist field site. In all cases, the mean transport time of the microspheres was shorter than the solutes, regardless of detection limit. In all but the smallest scale test, only a fraction of the injected microsphere mass was recovered, with the smaller microspheres being recovered to a greater extent than the larger microspheres. Using existing theory, we hypothesize that the observed microsphere early arrival was due to volume exclusion and attenuation was due to aggregation and/or settling during transport. In most tests, microspheres were detected using flow cytometry, which proved to be an excellent method of analysis. CML microspheres appear to be useful tracers for fractured rock in forced gradient and short‐term natural gradient tests, but longer residence times may result in small microsphere recoveries.