To explore the effect of biodiesel and sulfur content on PM2.5 emissions, engine dynamometer tests were performed on a Euro II engine to compare the PM2.5 emissions from four fuels: two petroleum diesel fuels with sulfur contents of 50 and 100 ppm respectively, and two B20 fuels in which soy methyl ester (SME) biodiesel was added to each of the above mentioned petroleum diesel fuels (v/v: 80%/20% for petroleum diesel and SME respectively). Gaseous pollutants and PM2.5 emissions were sampled with an AVL AMA4000 and Model 130 High-Flow Impactor (MSP Corp). Measurements were made of the PM2.5 mass, organic carbon (OC), elemental carbon (EC) and the water-soluble ion distribution. The results showed that PM2.5 emissions decreased with lower sulfur content or blending with SME biodiesel, and the decrease would be more by applying both two methods together. Particles of approximately 0.13 μm contributed 48–83% of PM2.5 emissions. The impact of sulfur content on this percentage was different for low and high engine speed. The majority of PM2.5 was comprised of OC and EC, and the carbon emission rate had the same trend as PM2.5. Since the EC abatement of B20 was larger than OC, the OC/EC ratio of B20 was always larger than that of petroleum diesel. For petroleum diesel, the OC/EC increased with sulfur content, which was not the case for B20. The SO42- had highest emission rate in the water-soluble ions of PM.