TORONTO - Drivers in southern Ontario could save millions of dollars by having their cars tested less often through the province's Drive Clean Program, with only a marginal increase in pollution, a new study suggests.

The University of Guelph study found the money being spent by drivers on the Drive Clean Program would be better spent helping curb pollution, said co-author John Livernois of the university's department of economics.

The Drive Clean Program targets vehicles in the so-called "smog zone'' of southern Ontario, which is the area of the province with the highest population density and greatest number of vehicles.

Drivers pay $35 every two years to have their vehicle tested to ensure it's in line with provincial standards on vehicle emissions. If not, repairs are mandatory to get the vehicle up to snuff.

Cars five years and older are tested every two years; those more than 20 years old are checked annually.

But testing fewer vehicles less often would cost just one-fifth of the current $131 million, though emission reductions would be only 70 per cent of what they are now, the study's authors said.

"By making some small changes ... the cost savings are very large and the loss in pollution reduction is very small,'' Livernois said.

To compensate for the increased emissions, a new fee could be assessed on drivers and funnelled into an alternative pollution reduction strategy such as better public transit, Livernois said.

A spokesman for the Drive Clean Program, which is run through Ontario's Ministry of the Environment, was blunt in his assessment of the study.

"I don't think we would want to do anything that would increase emissions,'' said Charles Ross. "What price do you put on bad air?''

Ross said the stringent program encourages motorists to keep their cars more finely tuned.

"You know your vehicle has to pass a drive clean test, so why fail and cause yourself extra grief?'' said Ross, adding there are benefits like better gas mileage and improved dependability when a car is kept in tip-top shape.

Ross said the program recently underwent a comprehensive, rigorous review, ending with a series of changes in January 2006 that target vehicles most likely to pollute.

In other words, it's been fixed.

"If we were not testing cars, we would not be identifying vehicles that fail the test,'' Ross said.

The study says testing is more effective if it begins when a vehicle is six or seven years old, instead of at five years, and when it is conducted less frequently.

That's because the failure rate for newer cars is very low, while repairs on older cars usually last longer than one year, the authors said.

About 10 per cent of the 2.5 million vehicles tested each year fail.

The three-year study is based on data from 1999 to 2001, which were the most recent available.