MOUNT LAUREL, N.J. -- At one time, testing high school athletes for steroids was seen as the best way to fight performance-enhancing drug use among the young. Now, those efforts are losing steam because of high costs and few positive results.

New Jersey, Florida, Texas and Illinois have tried steroid testing since 2006, and an examination of the results by The Associated Press shows that only 20 tests out of 30,799 have come back positive.

That's far short of what one study concluded about use of the drugs that are associated with stunted growth, hormonal problems, strokes and heart ailments. University of Michigan surveys conducted in 2007 and '08 each found 2.2 percent of seniors said they had tried steroids at least once -- down from 4 percent in 2002.

Testing advocates argue that results from the four states show the program works as a deterrent. Critics say they show the flaws in how the tests were conducted. Either way, it's becoming harder amid a recession to justify spending up to $200 each on tests that rarely catch cheaters.

Missouri state Sen. Matt Bartle tried to push his colleagues to adopt a statewide high school steroid testing program because he was concerned that young athletes were emulating the bad habits of some professionals.

But when Florida dropped its program in 2008 after a costly one-year trial in which there was only one positive out of 600 students tested, Bartle decided a similar effort wouldn't be cost-effective in Missouri, and he didn't submit a proposal this year.

"Is there enough steroid use out there that spending a couple million bucks a year against everything else that the state needs to spend money on is worth it?" Bartle asked.

The state programs grew out of health concerns and doping scandals in baseball, cycling and track and field. Last month, New York Yankees slugger Alex Rodriguez became the latest name tied to performance-enhancing drugs, admitting he used them while with the Texas Rangers from 2001-03.

All four states chose athletes at random. In New Jersey and Illinois, only those on teams in state tournaments were subject to the testing, while all athletes in Texas were, although no tests were given in the summer after the academic and athletic year is over. In Florida, all participants in six sports (baseball, football, softball, girls' flag football and boys' and girls' weightlifting) were eligible.

The AP's examination of the states' steroid tests showed limited impact when it came to catching users:

n While 20 tests came back positive, six were granted medical exemptions.

n Another 12 tests in Texas are being rechecked because the results were unresolved. Officials there also have classified another 70 cases as "process positives" because students refused to give urine samples or had unexcused absences when they were called.

n Accounting for the exemptions and process positives, and assuming every unresolved test is positive, that still means tests revealed 78 steroid users at most. That translates into one user for every 320 tests given.

Bob Colgate, assistant director of the National Federation of State High School Associations, used to get calls from sports administrators across the country asking about steroid testing. As state budgets have tightened, he said, such calls have ended. And those involved with high school sports still debate whether testing is effective.

"Did we have a problem? Do we have a problem, or is this taking care of the problem?" Colgate asked. "I don't think everybody's come to grips with this."

Such uncertainty doesn't play well in legislatures.

Critics in Texas say the program has failed to get enough positive results to justify the cost -- $6 million to test up to 50,000 students for two years.

State Sen. Dan Patrick of Houston said in 2008 that the program should be abolished, calling it "a colossal waste of taxpayer money." He was unsuccessful, although testing advocates concede the program is likely to be scaled back for the upcoming school year.

"I could support something that is narrowed and tailored to focus on the students and sports who are more likely to be implicated in those tests," Patrick said.

The state representative who championed the program counters that testing works as a deterrent.

"The idea was not a 'gotcha' program," said Texas state Rep. Dan Flynn. Because of testing, coaches, parents and students know more about dangers of steroids and how to identify the signs that a teen is using them, he said.

New Jersey state Senate President Richard Codey said he knew something should be done in 2005 when his basketball-playing sons -- then in high school and college -- told him they were aware of peers who used steroids.

At Codey's urging, the state began testing for the 2006-07 school year, randomly checking students who were in state tournaments. The cost of $100,000 per year is split between the state government and the New Jersey Interscholastic Athletic Association.

In the first two years, two of the 1,001 tests given to players chosen at random from playoff teams in all sports came back positive.

Linn Goldberg, a professor at the Oregon Health and Science University, said New Jersey's program -- which is similar to Illinois' -- doesn't work because students know they'll only be tested during the playoffs.

That means they can use steroids with no chance of being caught for most of the year, he said.

Frank Uryasz, president of Drug Free Sport of Kansas City, which conducts testing for all the states with mandatory tests, as well as for the NCAA, said surveys on steroid use by college athletes show that testing is a deterrent.

He acknowledges there are built-in problems in testing high school students.

"How likely is it we're going to test a high school athlete in July?" he asked. "Zero. It's not going to happen."

Codey disagrees.

"You've got to think it makes a kid say, 'I'm going to lay off for the next four months,'" he said. "That's a positive."

Goldberg, who developed a steroids education program that was implemented in schools with funding from the NFL, said testing is not "a quick fix." He added: "There has to be peer pressure to do the right thing."

Zach Greenwald, a star on the powerhouse football and wrestling teams at New Jersey's Paulsboro High, said peer pressure works.

He said he's not aware of anyone at his school using steroids and he isn't tempted because he knows about their health effects.

Still, the 17-year-old junior said not everyone his age has the same perspective, which is why he doesn't mind that he was tested last fall.

"It cuts back on teams cheating," he said.

n Associated Press writer Jim Vertuno in Austin, Texas, contributed to this story.