Characterized by a high frequency of use, harsh working environments, poor maintenance, and low levels of emission controls, rural vehicles (RVs) are becoming an important source of air pollution. Our study used a portable emission measurement system (PEMS) to test the real-world emissions of 35 RVs on provincial, rural, and farm roads. The results show that high emission rates of carbon monoxide (CO) and hydrocarbons (HC) mainly occurred when accelerating at low speeds. However, high levels of nitrogen oxides (NOx) were emitted during high-speed acceleration. The particulate number (PN) of emissions was higher when the RVs were accelerating. According to the overall test results, the vehicle specific power (VSP) on the provincial road mostly ranged within (0, 3], accounting for 68.80% of the total. The VSP on rural and farm roads was concentrated within (0, 2] kW·ton−1, accounting for 67.09% and 76.64% of the total, respectively. We defined 14 bins based on the distribution of the VSP values and calculated the average emission rate of each bin. By comparing the average emission rate among the bins, we found that within Bins 1–7 (VSP < 0 kW·ton−1), CO, HC, and NOx emissions slowly increased as the VSP increased. In Bins 8–13 (VSP ≥ 0 kW·ton−1), the average emission rates of four pollutants increased as the VSP increased. However, all pollutants decreased in Bin 14 (VSP ≥ 6 kW·ton−1). We built a microscopic emission model according to the VSP distribution characteristics of RVs on different road types. We compared the measured and simulated emission factors and found that our emission model can greatly simulate the HC, NOx and PN emission factors of RVs.