This work aims to develop statistical models for ultrafine/fine particle number emission rates from a diesel bus, to evaluate the explanatory power of engine operating variables. Emissions were recorded by using on-board instrumentation in two types of real-world driving conditions: a freeway commuting route and a within-city-limits bus route, with stop and go due to intersections and bus stops. To reduce the risk of drawing spurious conclusions, three replications of the experiment were performed and linear models were estimated using the robust-to-outliers quantile regression method. The set of explanatory covariates examined includes engine speed, engine load percentage, fuel to air ratio, injection pressure, boost pressure and exhaust temperature; statistical models were corrected for variations in ambient temperature. Ultrafine Particles, Quantile Regression, Statistical Model, Emission, Linear Model.