A second carbonyl round robin was conducted to enable participating laboratories doing routine analysis of carbonyls in vehicle exhaust emissions to assess their analytical capabilities. Three sets of solutions in acetonitrile containing varying number and amounts of standard DNPH-carbonyls were prepared. The parent carbonyls are known components of vehicle exhaust emissions. The samples were designed to challenge the capabilities of the participants to separate, identify and quantify all the components. The fourteen participating laboratories included automotive, contract, petroleum and regulatory organizations. All participants were able to separate and identify the C3 carbonyls; a few were not able to separate MEK from butyraldehyde and methacrolein from butyraldehyde; and many were not able to separate adequately the isomers of tolualdehyde. Inadequate separation and lack of appropriate standards resulted in a few misidentifications. All participating labs have excellent analytical systems quality control. The interlaboratory average of the %rsd's (percent relative standard deviation) of replicate measurements ranged from 0.2 to 3.8 for all of the analytes. The recoveries of components that were separated and unambiguously identified are excellent. Interlaboratory component variabilities are generally less than 10% rsd. Recoveries of 100±6% were achieved.