More than 50 pesticides tested by a discredited laboratory have been approved for "emergency" use on products from soybeans to fruit, despite doubts about their safety, a House Agriculture subcommittee was told yesterday.

Allen Spalt, a researcher for the non-profit Rural Advancement Fund in North Carolina, said that a computer analysis of Environmental Protection Agency records for 25 states showed that more than 40 percent of 2,089 emergency registrations approved in 1981 and 1982 involved chemicals tested by Industrial Bio-Test Laboratories (IBT) of Northbrook, Ill. A pesticide may be requested for emergency registration many times.

IBT was responsible for health- effects tests on more than 200 pesticides through the mid-1970s. The EPA has declared most of those tests invalid in the wake of allegations of shoddy lab practices and data falsification at IBT. Four executives of the defunct laboratory are on trial for criminal fraud. They have denied the charges.

Spalt told the House panel yesterday that IBT performed the safety tests for Furadan, Sencor and Paraquat, the products most widely requested under a section of the federal pesticide law permitting emergency registration for "special local needs." In the last two years, the EPA approved 361 requests for emergency use of those pesticides in the 25 states he studied, Spalt said.

Spalt said his study also found that more than 90 percent of the special registrations are sought by chemical manufacturers, not farmers or state officials. And even though the emergency registrations are supposed to be limited to small-scale applications for special pest problems, he said, most of the pesticides are being used across broad areas on such crops as soybeans, cotton and grain.

The use of such pesticides on major crops "is particularly alarming because it means that untested chemicals can be routinely used in larger amounts over vast acreage," he said.

The report was released at a hearing convened to check the EPA's progress on handling the IBT situation and to find out how the agency intends to prevent similar problems.

"Many industry and government representatives argue that the IBT problem and other findings suggesting widespread problems . . . are behind us," said Rep. George E. Brown Jr. (D-Calif.), chairman of the subcommittee. "Many other people, I among them, are not prepared to accept these assertions on faith."

Edwin L. Johnson, the EPA's director of pesticide programs, tried to reassure Brown, telling him that "the major uncertainties caused by invalid data have been brought under control or are well on their way to being resolved."

Only a "handful" of pesticides registered in this country were "wholly supported" by IBT data, Johnson said. He said the agency has nearly completed notifying manufacturers that they must replace the studies or lose their registrations.

But Spalt and other witnesses for environmental groups disagreed, contending that a requested study was not the same as a completed study and that the EPA is attempting "to minimize public anxiety."

Under questioning about the agency's current practices, Johnson conceded that his office has had "resource" problems. But he contended that the agency's laboratory audit program is back on track and making progress "in balance with the other things we have to do."

Some members remained skeptical. Rep. Pat Roberts (R-Kan.) noted that in six years the EPA had referred only four cases of fraudulent laboratory practices to the Justice Department for prosecution. One of those--the IBT case--has been accepted.

"My gosh, we probably indict more congressmen in six years than that," Roberts said.