With ten million American kids a year prescribed Paxil or other popular anti-depressants, the question of whether the drugs actually cause suicidal behavior in children is crucial.
Last summer, a link was revealed in data from Paxil's maker, GlaxoSmithKline.
"It has us worried," said an FDA internal document.
So the FDA assigned one of its respected medical officers, Andrew Mosholder, to look at studies of eight anti-depressants and thousands of troubled kids.
What he found added fuel to the fire.
According to documents obtained by CBS News, there were twice as many suicide-related events in children who took antidepressants as those who got only sugar pills.
Mosholder called the findings "difficult to dismiss." But the public never got to hear those results. As Mosholder prepared to report to this FDA panel last month, there was an incredible turn of events.
Sources claim Mosholder's FDA bosses intervened and "pressured him to change his conclusions" to make them "ambiguous and less definitive."
At the last minute, they took his report off the agenda entirely.
Word of Mosholder's findings got out to the media anyway. That's when FDA managers launched a criminal investigation to find which employees leaked the report. Even though, according to one source, "it's clearly information the public should have."
Earlier this month, the FDA issued a warning on anti-depressants and suicide, but watered it down by implying there was really no strong evidence -- making no mention of Mosholder's findings.
Patient advocate with the Alliance for Human Research Vera Hassner Sharav says top FDA officials are watching out for the drug industry, not consumers.
"What we want to know is why the FDA has been sitting on this evidence and pretending, now after all this time, there is no evidence," Sharav says.
Congress is asking the same questions. Did the FDA try to suppress a link between anti-depressants and suicidal behavior in kids, and did it try to intimidate employees who exposed the risks.