This study used a new emissions modeling tool, CT-EMFAC, to assess the sensitivity of mobile source air toxic (MSAT) emissions to changes in traffic volumes, speeds, and fleet composition. In addition, the choice of the speed calculation method, which is usually part of a postprocessing technique, was analyzed. The investigation employed a hypothetical 6.7-mi freeway segment located in southern California; activity data were derived from comparable real-world information obtained from the California Department of Transportation. Results show that emissions more than doubled in 2004 and increased by a factor of two to four in 2030 when traffic volumes increased 30% above base case conditions. The nonlinear shift in emissions was a function of decreased travel speeds and increased grams per mile emission rates that accompanied increased traffic volumes. Fleet composition (the proportion of trucks) was also shown to affect MSAT emissions, especially for diesel particulate matter and aldehydes. Under some scenarios, the choice of the speed calculation method had a greater effect on MSAT emissions than 26 years of fleet turnover. The analysis also showed that the type of speed calculation method used could result in large variations in MSAT emissions. Application of one speed calculation method resulted in calculated congested freeway speeds that did not fall below 10 mph. The speed calculation method chosen was highly influential in the 2030 case study when forecasted volumes reached levels at which speeds approached the 10 mph minimum allowed by the speed calculation method; this probably underpredicted emissions.