The objectives of this study were to: (1) evaluate a new low-cost approach for measuring on-road tailpipe emissions of highway vehicles; (2) investigate factors that affect the amount and variability of on-road emissions, using statistical methods; and (3) devise and demonstrate methods for designing and conducting observational experiments that realistically evaluate pollution prevention strategies for on-road vehicles. Portable instruments were used for measuring carbon monoxide (CO), nitric oxide (NO), and hydrocarbon (HC) emissions and vehicle activity (e.g., vehicle speed, engine parameters) on a second-by-second basis. Data collection, quality assurance, reduction, and analysis protocols were developed. Field data collection occurred in a pilot and an evaluation phase. In total, over 1,200 one-way trips were made with more than 20 vehicles, 4,000 vehicle-miles traveled, 160 hours of data, and 10 drivers. The pilot study was used to identify key factors influencing on-road emissions and as input to the design of the evaluation study. In the evaluation study, data were collected intensively with a small number of vehicles on two corridors before and after signal timing and coordination changes were implemented. For the first corridor, changes in signal timing and coordination did not result in a significant change in traffic flow or emissions. However, substantial reductions in emissions were estimated for uncongested versus congested traffic flow when comparing travel in the same direction at different times of day. For the second corridor, there were significant improvements in traffic flow and some reduction in emissions for three of the four time period and travel direction combinations evaluated. The impact of signal timing and coordination changes with respect to non-priority movements involving cross-streets was evaluated. For the first corridor, there was no statistically significant observed change in emissions for non-priority movements. For the second corridor, there typically was a decrease in average speed and an increase in emissions for non-priority movements; however, many of the observed changes were not statistically significant. The study also demonstrated other analysis methods, including: (a) macro-scale analysis of trip average emissions and traffic parameters; (b) micro-scale analysis of second-by-second emissions and vehicle operation; (c) mesoscale analysis of modal emission rates; and (d) spatial analysis of emissions at specific locations along the corridors. Both statistical and theoretical-based approaches were evaluated. The implications of the study results for pollution prevention strategies are discussed. Conclusions are presented regarding instrumentation, protocols, analysis techniques, and case study-specific findings. Recommendations are given regarding future applications of on-board measurements.