In two separate tests rapid sand filtration was evaluated as a means of removing the waterborne triactinomyxon actinospores (TAMs) of Myxobolus cerebralis, the causative agent of salmonid whirling disease, from contaminated water. In the first test we independently evaluated two sizes of sandblast sands: #4010 (effective size, (ES) = 250 μm) and #4060 (ES = 300 μm). The product number relates to the sand's size composition, and each had a different effective size. Effective size is defined by that size fraction at which only 10% of smaller particles remain. Rainbow trout Oncorhynchus mykiss were stocked into two separate systems with filters containing either one or the other size of sand, and TAMs were added to each system three times per week for all but the final 2 weeks of the 16-week test. A positive control (in which fish were exposed to the same number of TAMs without filtration) and a negative control (in which fish were not exposed to TAMs) were also included. Infection quantified by the pepsin-trypsin digest method showed no infection in fish from the two filter treatments. The average myxospore load among the positive controls was over 90,000 spores per fish head. In a final test, a nonreplicated comparison of filtration systems was conducted. The three systems used were as follows: (1) a drum filter in line with a UV filter, (2) a media filter followed by an ultraviolet (UV) unit, and (3) a sand filter containing three layers of sand. Water contaminated with TAMS was run through each filtration system, as well as through a positive control system, to raceways containing rainbow trout. Both of the dual-component commercial systems were 100% effective at preventing infection; the sand filter was 92% effective.