The installed performance of cooking exhaust fans was evaluated through residential field experiments conducted on a sample of 15 devices varying in design and other characteristics. The sample included two rear downdraft systems, two under-cabinet microwave over range units, three different installations of an under-cabinet model with grease screens across the bottom and no capture hood, two devices with grease screens covering the bottom of a large capture hood (one under-cabinet, one wall-mount chimney), four under-cabinet open hoods, and two open hoods with chimney mounts over islands. Performance assessment included measurement of airflow and sound levels across fan settings and experiments to quantify the contemporaneous capture efficiency for the exhaust generated by natural gas cooking burners. Capture efficiency is defined as the fraction of generated pollutants that are removed through the exhaust and thus not available for inhalation of household occupants. Capture efficiency was assessed for various configurations of burner use (e.g., single front, single back, combination of one front and one back, oven) and fan speed setting. Measured airflow rates were substantially lower than the levels noted in product literature for many of the units. This shortfall was observed for several units costing more than $1000. Capture efficiency varied widely (from < 5 percent to roughly 100 percent) across devices and across conditions for some devices. As expected, higher capture efficiencies were achieved with higher fan settings and the associated higher air flow rates. In most cases, capture efficiencies were substantially higher for rear burners than for front burners. The best and most consistent performance was observed for open hoods that covered all cooktop burners and operated at higher airflow rates. The lowest capture efficiencies were measured when a front burner was used with a rear backdraft system or with lowest fan setting for above-the-range systems that do not cover the front burners.