absorption of the IR and thus determine the pollutant concentrations. Effectively, we have taken a hack saw to the box, placed the IR light on one side of the road, the detector on the other, and allowed the car to drive through. The system has been tested since 1987 and at speeds between 2 and 152 miles per hour. In 1991 in California, the CO readings were found independently to be within 95% of correct, while the HC were within 85% (Lawson and Gunderson, 1991 Report to California I/M Review Committee).

We measured passing cars on southbound Rosemead Boulevard in El Monte, Calif., in a roadside pull-over program. When we identified an apparent gross polluter, we radioed ahead for a California Highway Patrol officer to pull it over for a roadside Smog-Check test. Two teams of testers were used. The USEPA also had a portable dynamometer (treadmill) set up for a test called IM240 (now familiar to Coloradans) in the roadside park. In 10 days, the remote sensor was able to take 60,487 emission readings from 58,063 vehicles. Two SMog-Check teams were able to measure 340 while one IM240 system measured only 80. Although we are not professional economists,