WASHINGTON (AP) _ Upscale prostitutes are less likely to be carriers of the AIDS virus than are their inner-city colleagues, whose high rate of infection may be due to drug abuse, researchers said Wednesday.

Once infected, however, the prostitutes can easily transmit the disease to their customers, according to studies presented at the Third International Conference on AIDS.

A study by the federal Centers for Disease Control in Atlanta found that 56 percent of prostitutes in northern New Jersey cities, where drug abuse is common, were infected with AIDS, while rates of infection were much lower in prostitutes in areas with little drug abuse.

''The major risk factor (for AIDS in prostitutes) appeared to be intravenous drug abuse, not sexual exposure with infected partners,'' said William Darrow, director of the Centers for Disease Control study.

The study comparing inner-city prostitutes with a middle-class escort service in Miami found that 37, or 41 percent, of 90 inner-city prostitutes were infected with AIDS. But none of 25 prostitutes working for the escort service were infected, said the study's director, Margaret Fischl of the University of Miami.

The inner-city prostitutes were more likely than the others to share hypodermic needles, a known means of spreading AIDS, and they were also more likely to have contact with AIDS-infected customers, Fischl said.

In addition to drug-related infections, ''we suspect we were seeing heterosexual transmission,'' Fischl said. Twenty-seven of the 90 inner-city prostitutes did not use intravenous drugs and eight of them were infected, she said.

In other developments:

-Deaths from tuberculosis, heart infections and pneumonia among drug addicts infected with the AIDS virus have risen dramatically since the AIDS epidemic began in 1987, said Dr. Don. C. Des Jarlais of the New York State Division of Substance Abuse Services. Des Jarlais said that as many addicts may be dying of the AIDS-triggered illnesses are being killed by AIDS itself.

-Costs of treating AIDS patients in the United States could exceed $37 billion from mid-1986 to mid-1991 and could reach $113 billion if there is a large increase in the number of patients and in the cost of medical care, the Rand Corp. said Wednesday. The study includes only costs of hospital and outpatient medical care and not costs of social services or earnings lost due to illness and premature death.

-An experimental AIDS vaccine has given a modest boost to the immune systems of 10 AIDS victims and 12 healthy people who have received it, according to Daniel Zagury, a University of Paris researcher who has injected himself with the vaccine. The vaccine, which containes fragments of the virus that causes acquired immune deficiency syndrome, has not caused ill effects among the recipients, but Zagury said it was still to early to tell if the vaccine is effective in preventing AIDS infection in healthy volunteers.

-A study by the American Red Cross in Los Angeles estimates that chances are one in 64,000 that a person could be infected with the AIDS virus by receiving a transfustion of tainted blood that was not detected by tests designed to screen out AIDS.

-Great Britain's effort to educate people about AIDS, which included distributing leaflets to each of the nation's 23 million households, was ineffective, according to a pair of London researchers. The study found that the public education effort resulted in ''slight but significant knowledge gains'' and would have benefitted from better planning.

Darrow's study of prostitutes in seven areas around the country found widely divergent rates of infection, including no infection in a group of 26 prostitutes in Las Vegas.

He also found that AIDS was much more common in black and Hispanic prostitutes than in white prostitutes, and that condoms were an effective protection against AIDS. Only one of 10 women who always required customers to use condoms became infected, and she was a drug abuser, Darrow said.

Some authorities have speculated that AIDS transmission during heterosexual contact might be easier in people suffering from sores or lesions associated with other sexually transmitted diseases.

In order to investigate that, James Goedert of the National Cancer Institute studied the spread of AIDS from healthy, middle-class hemophiliacs to their partners. The hemophiliacs acquired AIDS through blood transfusions and had no evidence of other sexually transmitted diseases.

Goedert found that ''transmission of HIV through heterosexual contact occurs readily.'' HIV, or human immunodeficiency virus, is the cause of AIDS.

Among 18 couples who did not use condoms, four of the hemophiliacs' spouses became infected. In six couples who abstained from sexual contact or always used condoms, no transmission of the disease occurred, Goedert said.

Other studies have shown low rates of infection of partners of hemophiliacs, but Goedert found that hemophiliacs become more infectious as their disease worsens, as measured by a drop in the number of so-called T4 cells, which are attacked and killed by the AIDS virus.

Goedert said that those hemophiliacs who had not infected their partners might be about to do so in the near future, as their T4-cell counts drop.

''Testing for T-cell subsets not only assists in predicting the risk of AIDS but also predicting the risk of infection,'' Goedert said.

Neal Steigbigel of the Albert Einstein College of Medicine in New York City, who has also studied heterosexual transmission of AIDS, said that the studies presented ''give no evidence whatsoever on the risk of a single encounter.''

For susceptible individuals, a few sexual contacts might be enough to acquire the disease; for others, regular sexual contact over an extended period might not allow the disease to spread, he said.