Vehicle self-pollution occurs when a vehicle's emissions migrate to inside that vehicle's passenger compartment. This paper presents values for two new parameters: vehicle self-pollution intake fraction (iFSP), which is the total fraction of a vehicle's emissions inhaled by all people in the vehicle, and vehicle self-pollution individual intake fraction (iF i,SP), which is the fraction of a vehicle's emissions inhaled by an individual in the vehicle. We use results from tracer-gas experiments in California's South Coast Air Basin (SoCAB) to quantify students' iFSP and iFi,SP for school bus emissions. Six buses were studied during nine runs with windows open and seven runs with windows closed. Reported iF SP values (units: per million; min = 10, max = 94, mean = 27) indicate that the total mass of a bus' exhaust inhaled by students commuting on it is comparable in magnitude to the total mass of that bus' exhaust inhaled by all other people in the SoCAB. Reported iFi,SP values (units: per million; min = 0.2, max = 2.4, mean = 0.7) indicate that average per capita inhalation of emissions from any single bus is 105-106 times greater for a student on that school bus than for a typical resident in the SoCAB. Vehicle self-pollution rate varies with bus window position (open or closed) and bus manufacture year. Our results can be used to develop cost-effective strategies to reduce children's exposure to school bus emissions. Our results indicate, for example, that even if emission reductions were many times more expensive per gram emitted for school buses than for an average vehicle, it would still be less expensive per gram inhaled by a student to reduce emissions from school buses than from an average vehicle.