The Emissions Research Laboratory (ERL) of the Arizona Department of Environmental Quality (ADEQ) has developed tests to determine, in situ, both the oxidation and reduction capabilities of light duty, spark ignition vehicle catalysts. The tests require minimal system intrusion; determination of oxidation efficiency is essentially nonintrusive, but reduction efficiency measurement requires replacement of the oxygen sensor with a high precision universal exhaust gas oxygen (UEGO) sensor.The vehicle is operated on a dynamometer at 40mph steady state using the IM240 specified loading for that vehicle. To ensure proper fueling, the engine-out A/F ratio is measured. This measurement can be taken at the tailpipe with secondary air injection disabled, if present. Oxidation efficiencies for hydrocarbon (HC) and carbon monoxide (CO) require normal vehicle configuration (secondary air active, if present) and tailpipe measurements of total HC and methane (CH4). The catalyst efficiencies for oxidation of HC and CO are inversely related to the HC/CH4 ratio. The reduction efficiency for nitrogen oxides (NOx) requires NOx tailpipe measurement at two A/F ratios, stoiciometric and about 15/1 (wt/wt). At about 15/1 A/F, excess oxygen prevents catalyst reduction and, consequently, an engine-out value can be obtained from a tailpipe measurement. The difference between the tailpipe measurements at the two A/F ratios provides an experimental catalyst efficiency which can be related, by regression analysis, to the actual efficiency.Actual catalyst efficiencies for HC, CO, and NOx were obtained from concurrent engine-out and tailpipe emissions measurements. Data were obtained from 27 vehicles using the original catalyst and (primarily) two intermediately active catalysts. Actual efficiency data show that catalysts with HC/CH4 ratios less than 3.5 will, with a high degree of confidence, have oxidation efficiencies above 80% for HC and CO, and NOx reduction efficiency can be determined at a standard error of estimate of 6%. Engine-out emissions differences (for different vehicles and at different A/F ratios) increases data scatter as catalyst efficiency decreases. NOx differences do not excessively influence test results for catalysts having >10% reduction capability, but at oxidation capability below 80%, HC and CO engine-out differences produce data scatter to the extent that the primary significance of the test is determination of maximum or minimum oxidation efficiency.