statistical study of more than 300,000 car loans arranged through Nissan dealers from March 1993 to last September — believed by experts to be the largest pool of car loan data ever analyzed for racial patterns — shows that black customers in 33 states consistently paid more than white customers, regardless of their credit histories.

The gap between black and white borrowers was largest in Maryland and Wisconsin, where the average finance charge paid by blacks was about $800 higher than whites paid.

Among the largest states, the study also showed that on average blacks paid $245 more in Connecticut, $339 more in New Jersey and $405 more in New York. In Texas, the black-white gap was put at $364; in Florida, it was $533.

"Unfortunately, since we have not been able to examine this new material, we cannot comment yet on these new studies," said Dierdre Dickerson, a spokeswoman for the Nissan Motor loan unit, Nissan Motor Acceptance Corporation.

"But NMAC has never tolerated unfair treatment of any of its customers, let alone racial discrimination," Ms. Dickerson said.

Nissan, like other auto lenders, lets car dealers make the final decision about how much customers pay for loans. But class-action lawsuits filed in Nashville against the Nissan loan unit and the General Motors Acceptance Corporation, as well as similar suits against other auto lenders, argue that the lenders are responsible if dealers set rates in a racially disparate way.

None of the companies are accused of racial bias — indeed, both sides agree that the giant lenders do not even know the race of the customers whose cars they finance.

Rather, the plaintiffs argue that the lenders' credit policies, for whatever reason, are having a statistically disparate impact on blacks. If they can prove that, the legal burden shifts to the lenders, who must then show a valid business need for the policies, which cannot be met in a less harmful way.

Indeed, one expert who examined the new data for the plaintiffs in the Nissan case suggests that some, though not all, of the racial disparity in loan charges may be an unintended result of limits NMAC puts on certain kinds of loans — loans that go disproportionately to white borrowers. Other contributing factors, he said, may be simple racial bias among dealers or an assumption among dealers that blacks will be less likely than whites to balk at higher credit prices.

The plaintiffs for whom these new studies were done are challenging the industrywide practice of letting car dealers rather than lenders make the final decision about the interest rate customers will be charged on car loans arranged through their dealer. Similar fair- lending cases have been filed since October by black consumers against the financing arms of Ford, DaimlerChrysler and Toyota.

Lenders like NMAC allow dealers to add several percentage points, known as the dealer markup, to whatever interest rate Nissan has established for customers based on their income and credit history. That becomes the rate quoted to the customer, who has no way of learning of the disparity. The lender then splits the money generated by those extra percentage points with the dealer.

Dealer markup is a big-ticket issue for the industry, and the new Nissan data show why. One of the studies, done for the plaintiffs by a Vanderbilt professor, Mark A. Cohen, estimated that Nissan's share of the dealer markup paid by black consumers alone since 1990 was more than $210 million. Based on fee-splitting formulas in place during much of that time, the dealers' aggregate share would have been roughly three times that amount.

The Nashville cases were originally based on a smaller pool of loans in Tennessee. Based on the new data on Nissan — the only lender that has provided such information so far in the cases — lawyers for the black consumers are now seeking a preliminary injunction that would prevent Nissan from continuing to participate in dealer-markup arrangements nationwide.

NMAC has argued that any action should be postponed until a decision is reached in its appeal of an order certifying the class action.

Auto finance companies say that the dealer-markup system, besides generating revenue, allows lenders to compete for fresh business from car dealers, who can steer a customer's loan application to any of a number of financing outlets. Dealers say their share of the markup compensates them for providing customers with one-stop shopping and access to a broader spectrum of lenders than the consumer could command alone.

And the lenders themselves have argued in court that they are not responsible if the markup system is being applied in a racially discriminatory way by some car dealers.

That argument has been challenged by the Justice Department, which last fall filed a brief for the plaintiffs arguing that federal fair- lending laws hold lenders responsible for the actions of the dealers to whom they grant discretion.

The Nashville lawsuits were filed almost three years ago, but were first disclosed publicly last October by The New York Times and ABC's "20/20," which had jointly petitioned the Federal District Court there to lift the court orders sealing the files in both cases.

A spokeswoman declined to comment on the case pending in Los Angeles against Toyota. Dan Jarvis, a spokesman for Ford Motor Credit, which has been sued in Manhattan, said Ford had "zero tolerance for any form of discrimination." Chrysler Financial did not respond to questions about the case pending against it in New Jersey.

As the most mature of the pending lawsuits, the NMAC case remains the bellwether in the battle over the dealer markup system. And the new studies pose a roster of fresh questions about the practice.

The first study, by Professor Cohen, showed that blacks were more than twice as likely as whites to be charged a dealer markup in the first place. In the national Nissan data, 72.8 percent of blacks were charged a markup, compared to 46.7 percent of white customers.

And among all customers who were charged a markup, blacks consistently paid more than whites, according to Professor Cohen. Black borrowers who paid a dealer markup were charged, on average, $1,351, compared to $989 for the whites whose loans were marked up.

Moreover, the Cohen study showed that dealers were far less likely to give blacks, even those with top-quality credit, the advantage of the lender's preferential interest rates. Of the customers in NMAC's first credit tier, for example, 32.4 percent of whites were offered special interest rates, while only 15.3 percent of blacks were.

Professor Cohen tested for a number of variables that the lenders' statistical experts had offered as possible explanations for the observed racial disparities in Tennessee. None, he said, were sufficient to explain away all the differences in markup charges.

Professor Cohen cites, as an example, the case of two Louisiana women who purchased Nissan Sentras last year. Both women were in NMAC's first credit tier and were thus qualified to borrow at 8.25 percent. The first, who was white, borrowed $15,093 over five years at the preferential rate of 3.9 percent, with monthly payments of $277.73. The second, who is black, borrowed only $14,787 over five years, but she was charged a dealer markup of one percentage point, for a total interest rate of 9.25 percent. Her monthly payment was $309.94.

Nissan also permits dealers to move approved customers from a higher credit classification to a lower one, which has the effect of raising the customer's interest rate. Professor Cohen found that 4 percent of the black customers had been moved into a lower credit tier, while only 2.3 percent of white customers were.

NMAC established a limit on how much dealer markup can be imposed for each credit tier — and policies like that may be inadvertently contributing to the racial disparity that shows up in the loan data, according to Ian Ayres, a lawyer and economist who is a professor at Yale Law School.

Professor Ayres, who has conducted several widely cited studies on car retailing over the past decade, examined the new Nissan data for the plaintiffs in the NMAC case.

He noted that NMAC limited the amount of markup allowed on loans in the highest credit tier. Because whites are more likely than blacks to be placed in that tier, that limitation has the side effect of reducing the markup paid by whites. Similarly, NMAC allows greater markup on loans of longer duration. Because blacks tend to borrow for a slightly longer term than whites, this too can contribute to the higher markups they pay.

Finally, NMAC often imposed lower markup caps on loans for new cars, which whites are more likely to buy than blacks.

But Professor Ayres noted that there was no business justification for allowing dealers, who have no money at risk in the transaction, to impose a higher markup based on factors such as the credit tier, the loan term or the age of the car.

"None of these selective markup caps are justified by the dealerships' costs of arranging and or processing loan applications," Professor Ayres noted in his study.