The effect of lubricant composition on vehicle exhaust emissions has been investigated. Emissions from two vehicles were measured when lubricated with four different crankcase lubricants. All emissions tests were performed with California Phase II gasoline over the FTP-75 cycle. The lubricants tested were a conventional mineral oil based lubricant, poly-alpha olefin (PAO) based lubricant, hydrocracked based lubricant and a Volvo first fill lubricant. The first three lubricants were designed to have similar high temperature viscosities whilst using the same additive package. This meant that there were some small differences in the low temperature viscosities. This resulted in the two mineral oil based lubricants being 10W-30 grades and the PAO and hydrocracked based lubricants being 5W-30 grades.The two test vehicles used were both Volvo 850 vehicles, however one was a European specification vehicle and the other a Californian TLEV. The European vehicle was equipped with a five cylinder 2.0 litre engine. The TLEV was powered by a five cylinder 2.5 litre engine, and was fitted with a secondary air pump and small volume starter catalyst immediately upstream of the main catalyst. This additional hardware enabled the TLEV to achieve rapid catalyst light off. Each vehicle/lubricant combination was tested a minimum of four times.Lubricant composition had no effect on the emissions of total hydrocarbons, non-methane hydrocarbon, non-methane organic gases, speciated hydrocarbons or carbon monoxide. NOx emissions from both vehicles were affected by lubricant composition, with the NOx emissions from the TLEV being more sensitive to compositional changes in the lubricant. The PAO and hydrocracked lubricants gave lower NOx emissions, by up to 29%, compared to the mineral oil based products. These reductions can be explained by small differences in the tailpipe mixture strengths.