Engine-out and tailpipe exhaust, and hot soak evaporative emissions of two reformulated test gasolines and an Industry Average reference gasoline were compared in four vehicle fleets designed for progressively lower emission standards. The two reformulated gasolines included: 1) a gasoline meeting 1996 California Phase 2 regulatory requirements, and 2) a gasoline blended to the same specifications but without an oxygenated component. These two gasolines were compared with the Auto-Oil Air Quality Improvement Research Program's (AQIRP) Industry Average gasoline representing 1988 national average composition. The vehicle fleets were the AQIRP Older (1983 to 85MY) and Current (1989MY) vehicle fleets used in prior studies, and two new AQIRP test fleets, one designed to 1994 Federal Tier 1 standards and a prototype Advanced Technology fleet designed for lower emission levels of 1995 and later.The California Phase 2 gasoline (C2) was compared to the Industry Average gasoline (A) in the Older, Current, and Federal Tier 1 fleets. NMHC emissions were 12 to 27% lower with reformulated gasoline C2 than with gasoline A. Carbon monoxide was 21 to 28% lower, and NOx was 7 to 16% lower. All of these differences were highly or marginally statistically significant according to AQIRP criteria.Ozone-forming reactivity, expressed as reactivity-weighted emissions and specific reactivity, was calculated using 1991 SAPRC MIR factors. Reactivity-weighted emissions (RWE) were significantly lower by 16 to 30% with gasoline C2 vs. gasoline A in all three fleets. Specific reactivity of the tailpipe emissions varied from 3 to 9% lower with gasoline C2 across the fleets. The 3% reduction in the Federal Tier I fleet was marginally significant. The effects in the other two fleets were significant.Total toxic air pollutant emissions were 8 to 32% lower with gasoline C2 than with gasoline A. These differences were statistically significant and represent the net effect of changes in the two predominant species, benzene and formaldehyde, and two minor species, 1,3-butadiene and acetaldehyde. Significant decreases in benzene outweighed smaller and not always significant increases in formaldehyde.Differences between reformulated test gasoline C2 with oxygenate (MTBE) and similar gasoline C1 without oxygenate were generally not statistically significant. These gasolines were tested in the Current, Federal Tier 1, and Advanced Technology fleets. The Current fleet showed the largest differences, and the results were generally consistent with earlier Auto/Oil results found to be significant when testing was conducted in a fleet twice as large. The only significant effect in this study was a 13% decrease in formaldehyde emissions from the Advanced fleet when the fuel had no oxygenate.Comparison of tailpipe emissions from 4-cylinder passenger cars in the various fleets when using fuel C2 showed progressively lower NMHC in later model years, lower CO and NOx emissions in the newest vehicles, and lowest NMHC, CO, and NOx in the Advanced Technology vehicles. These reductions are attributed mainly to advances in catalyst technology and better fuel control in the modern cars.