For catalytic converter-equipped light-duty gasoline vehicles (LDGV), the hot-stabilized tailpipe emissions for pollutants such as carbon monoxide (CO), hydrocarbons (HC), and nitrogen oxides (NOx) are well controlled. However, there are few reported real-world measurements of cold starts. Idling cold start and hot-stabilized trip exhaust emissions were measured for 37 LDGVs using a portable emissions measurement system (PEMS). Five vehicles were also measured for transient driving cold starts. On average, it took approximately 400, 150, 330, and 120 s to accumulate 90 percent of the idle cold start increments for fuel use, CO, HC, and NOx, respectively. Driving cold start increments were substantially higher than idling cold start increments, whereas cold start duration was typically shorter. For example, driving cold start contributed approximately 64%, 68%, 58%, and 4.5% of the trip total CO, HC, NOx, and carbon dioxide (CO2), respectively. This study is unique in quantifying the cold start contribution on a trip basis with real-world data. Although the cold start increment is sensitive to driving compared to idling, in either case, cold starts contribute substantially to total exhaust mass emissions. Furthermore, driver decisions regarding driving versus idle can substantially affect the contribution of cold starts, especially for CO and NOx. [ABSTRACT FROM AUTHOR]