Lester Friedlander, now a consumer advocate, was fired from his job as head of inspections at a large meat-packing plant in Philadelphia in 1995 after criticizing what he called unsafe practices.

Friedlander said he knows U.S. Agriculture Department veterinarians who sent suspect cow brains to private laboratories that confirmed mad cow infection, but samples from the same animals were cleared by government labs.

"It's several veterinarians that have given me similar stories about sending cow brains in,'' he said in an interview Tuesday. "It might be shocking for Canadians but it wouldn't be shocking for veterinarians that have worked for the USDA.

"I'm willing to back this up with a voice stress analysis test or even a lie detector test.''

Friedlander wouldn't name the veterinarians, saying they still work for the Agriculture Department and would be fired if identified.

The department has denied Friedlander's allegations, which were first reported last week.

Rob McNabb, a spokesman for the Canadian Cattlemen's Association, said it does seem puzzling that four mad cow cases have been detected in Canadian-born cattle but none in U.S.-born cattle.

"It's true that the risk ... is very similar, and it is surprising,'' he said.

There are 120 million cattle in the United States, 15 million in Canada.

"I guess there's always going to be people raising the question, `How come it's 4-0?','' he said.

But McNabb wouldn't comment on Friedlander's allegations.

Michael Hansen, a scientist with the U.S. Consumers Union in Washington, said there's widespread suspicion about the testing of three suspected cases of mad cow in U.S. cattle.

Hansen said all tests came back negative in the three cases but the USDA used a rapid test based on immuno-histochemistry, not the Western blot test which is considered most reliable.

"Many of the top scientists think that's insane,'' he said of the use of the less reliable test.

He said there are also suspicions about a recent case in St. Angelo, Tex., when officials at an abattoir noticed a cow was staggering and wanted it tested, but permission was refused.

"The federal inspectors and the plant employees all wanted to test the animal and basically (the USDA) said, `Nah, we're not going to do that.' So the animal was sent to rendering and was never tested.''

Hansen said there appears to be a great lack of eagerness to detect mad cow in the United States.

A study by the Harvard Center for Risk Analysis three years ago concluded there was a 20 per cent chance that mad cow was present in the United States.

The U.S. government closed its border to live cattle imports from Canada in 2003 after a single Canadian cow tested positive for the disease. Three other Canadian cases have been confirmed since then, one in a Washington State cow that originally came from Canada.

The border was to reopen to live cattle March 7 this year but that was delayed by a challenge from a U.S. cattle industry lobby group.

Friedlander was in Ottawa to testify at a Commons committee examining proposed changes to the Canadian food regulation system.