A vehicle emissions round robin test program was conducted using a methanol-fueled vehicle operating on M85. Each of 16 participants conducted two to six Federal Test Procedure (FTP) emissions tests. All participants measured emission rates of hydrocarbon (HC), CO, NOx, methanol, formaldehyde, and acetaldehyde. One participant, designated as a reference lab, conducted emissions testing at the beginning, the end, and two intermediate times during the round robin. Results of the reference lab demonstrated that no significant drift in emissions levels occurred during the 2-year program.Relative lab-to-lab variability for FTP-composite emissions was lowest for NOx, with a coefficient of variation (C.V.) of 12%. CO variability was 16%, HC variabilities (by GC and bench FID) were 17 and 35%, respectively. Methanol, formaldehyde, and acetaldehyde were found to have variabilities of 34, 17, and 63%, respectively. The large difference between formaldehyde and acetaldehyde variability is probably a consequence of their different concentrations--formaldehyde's emission rate was approximately 40 times that of acetaldehyde. This is consistent with an expected increase in variability at emission levels near detection limits. Emission rates of three HC species were also determined: methane, benzene, and toluene. They showed variabilities ranging from 20-30%, only slightly higher than for the criteria emissions.