CIA documents released this week contradict testimony given a Senate committee by intelligence director Stansfield Turner and another intelligence agency official that no human subjects were involved in the final series of the agency's massive drug testing program.

Turner and Edward Gordon, who was described by the Central Intelligence Agency director as an "expert" on Project OFTEN, the last known CIA drug testing program, testified in September 1977 at the Senate subcommittee on health and scientific research that no humans were drugged as part of the project.

OFTEN was the last in a series of drug testing and mind control experiments on which the CIA spent millions of dollars. The experiments, kept secret for a quarter-century after they began in 1949, ultimately involved a range of sites stretching from San Francisco whorehouses to prestigious U.S. universities.

According to the latest documents, released under a Freedon of Information Act request filed before Turner's Senate testimony on Project OFTEN, the final program ran from 1967 into 1973. It was finally disbanded after newspaper reports of covert drug and chemical testing on humans by the CIA and the military.

The OFTEN project was originally designed to include drug testing on human subjects. But Turner and Gordon told Sen. Edward M. Kennedy (D-Mass.), chairman of the Senate subcommittee, that there was no evidence in the agency's files to indicate any humans were actually used in experiments before the project was terminated.

The documents state, however, that the CIA allocated $37,000 in 1971 to test a glycolate class of chemical on numans at the Army's Edgewood Arsenal Research Laboratory.

One report, compiled in 1975 under the heading, "Influencing Human Behavior," notes that 20 persons were tested with the drug -- five prisoners from Holmesburg State Prison in Holmesburg, Pa., and 15 Army volunteers.

The report does not say what the test results were except that side effects were evident up to six weeks later. The report notes that the CIA spent money for medical follow-up testing of the research subjects.

A CIA spokesman noted yesterday that while several memos and reports mention the testing, one 1975 memo states that an unnamed doctor at Edgewood told a CIA interviewer that year that no human subjects were used in the tests.

The spokesman noted that the program was set up to include human drug testing. "Maybe it did happen," he said, "but it's all academic now."