The preliminary findings are based on a review of 58 event data recorders — which capture information such as velocity, braking and acceleration — from vehicles that were in accidents blamed on unintended acceleration. In a majority of the cases, data showed that brakes were not applied before impact.
But the reliability of black box data has been questioned by automotive experts and even Toyota itself, and the officials were careful to point out that it is only a piece of the puzzle when it comes to analyzing sudden acceleration.
Transportation Secretary Ray LaHood and National Highway Traffic Safety Administration head David Strickland told three members of Congress in the briefing that the agencies had drawn "no conclusion" on the causes of sudden acceleration and noted that investigations by NHTSA, NASA, and the National Academy of Sciences were ongoing, with final results a year or more away.
"NHTSA and NASA are continuing to study whether there are potential electronic or software defects in these vehicles," the regulators told the congressmen.
Since last fall, Toyota has issued more than 11 million recall notices to address safety and quality problems, many of which are related to sudden acceleration. In April, the automaker agreed to pay a record $16.4-million fine for delaying a recall of sticking pedals.
The automaker is also facing hundreds of lawsuits, many of which allege that electronics or software are a potential cause of sudden acceleration, which has been blamed in at least 90 deaths, according to NHTSA.
A spokesman for Toyota said that the automaker had not yet reviewed the data presented to Congress.
But in a statement, the company said that its own reviews of some 4,000 vehicles had confirmed sticking pedals, floor mat entrapment, pedal misapplication and "vehicle functions where a slight increase in engine speed is normal" as causes of unintended acceleration.
"In no case have we found electronic throttle controls to be a cause of unintended acceleration," the statement said.
In recent weeks, black box data has been held up as a potential key to revealing the roots of the phenomenon, with Toyota publicly noting that black box data it had reviewed suggested drivers rather than vehicle defects were to blame.
As a result, three members of Congress wrote a letter to the Transportation Department several weeks ago requesting data on black boxes, said Rep. Michael C. Burgess (R- Texas), after unconfirmed media reports that the agency had discovered driver error in some cases.
In the briefing to Burgess and two other members of the House Energy and Commerce Committee, Reps. Bart Stupak (D-Mich.) and Ed Whitfield (R-Ky.), the regulators reported that among the 58 event data recorders, 35 indicated that drivers did not apply the brakes before impact.
That suggests that the accelerator was accidentally depressed rather than the brake pedal, mirroring claims by Toyota that in many cases sudden acceleration is the result of what it calls "pedal misapplication" or driver error.
But event data recorders, or EDRs, have also been criticized for being unreliable, with several academic studies calling their efficacy into question. Critics say that a software flaw, for example, could provide bad data to the EDR.
"We would not expect an investigation of the EDRs to find a problem with Toyota's electronics systems," said B. Craig Hutson, a debt analyst at Gimme Credit. "The EDRs are not designed to identify these types of problems. An electronics problem likely lurks in the millions of lines of software code found in a typical vehicle."
Toyota itself has repeatedly told state and federal judges that the readouts from the devices were too inaccurate to use in court, The Times reported last month.
In 2007, for example, a Toyota official told a New Jersey court that the automaker "cannot vouch for the reliability of any data downloaded" from EDRs.
Five of the 58 EDRs reviewed by NHTSA did not register any data at all; another contained data from a previous incident but not the crash blamed on sudden acceleration; and another contained what the agency termed "inconclusive data."
Meanwhile, 14 of the EDRs studied by regulators in fact did indicate that brakes were depressed at least partially before crashing, another suggested pedal entrapment and a final one revealed data that both the accelerator pedal and brake were applied at the same time.
Transport Canada, NHTSA's counterpart, said last week that its review of black box data did not indicate pedal misapplication.
According to a document given to Congress at Tuesday's briefing, NHTSA restricted its EDR review to 2007 or later model year Toyota and Lexus vehicles; to vehicles that crashed; and largely to incidents that occurred after mid-March of this year, when Toyota supplied the agency with 10 devices used to read EDR data.
Complaints about sudden acceleration in Toyota vehicles began spiking in 2001; by limiting the scope, NHTSA excluded some 4,500 complaints from review because they were for older models containing EDRs that did not capture as much data as more recent vehicles.
The document indicated that NHTSA recently conducted several controlled crash tests using Toyota EDR readers as a means of verifying the accuracy of the technology. It also showed that collecting data from black boxes was one small part of its vehicle inspection process, which includes such things as testing pedal function and interviewing witnesses.
The briefing document, which lists exactly what NHTSA does to test each vehicle, did not include testing the software or the electronics other than downloading the EDR data and looking for what are commonly known as check-engine codes.
Burgess said that the new findings were "a big piece of the puzzle." But he said they did not explain why complaints of sudden acceleration soared after Toyota's adoption of electronic throttle control beginning in 2001.
"The big question is whether electronic throttle control plays a role in this," Burgess said.