The University of Denver's remote sensor for on-road motor vehicle car
bon monoxide (CO) and hydrocarbon (HC) emissions was used in August an
d October, 1991, at two different locations in the Denver area. The re
sults of analysis for HC emissions show a similar skewed distribution
for the two data sets, although the average HC emissions are very diff
erent for the two data sets because different driving modes (no-load/l
oad) existed at the locations where the data sets were collected. Ther
e was no significant difference between HC emissions from those vehicl
es from counties in Colorado with and without an annual emission testi
ng program. The fraction of high emission polluters increases with veh
icle age. The evidence suggests that proper maintenance and eliminatio
n of tampering are very important even for those new vehicles which ar
e equipped with the latest emission control technology.