One approach to the treatment of urban runoff is to treat the runoff from critical source areas before it mixes, with runoff from less pollutant areas. Some of the general features of critical source areas appear to be large paved areas, heavy vehicular traffic, and/or exposed heavy equipment, materials or products. The control of runoff from relatively small critical source areas (such as loading docks, fueling areas, small maintenance yards, etc.) may be the most cost effective approach for the treatment/reduction of stormwater toxicants. However, in order for a treatment device to be usable, it must be inexpensive, both to purchase and to maintain, and be effective. Upflow filtration of stormwater was tested during both controlled tests, and under actual rainfall conditions, during SBIR1 (Small Business Innovative Research) and SBIR2 research funded by the US EPA. This paper summarizes the work presented by Pitt, et al. and Khambhammettu. Upflow filtration was originally developed to overcome some of the problems associated with conventional filtration. The most serious problem is that downflow filters clog relatively quickly, reducing the treatment flow rate potential and total treatment capacity, potentially causing large amounts of the stormwater to bypass the treatment units. Clogging does not occur as fast with upflow filtration. One reason is that the heavier particles get drawn away from the filtration interface due to gravity and fall into the sump which is an integral part of the upflow filter design.