A split Federal Trade Commission yesterday ended an eight-year effort to force food advertisers to document nutrition and health claims by killing a proposed rule that had been approved tentatively in 1980.

The commission's 2-to-2 vote reflected a clear change in enforcement emphasis since James C. Miller III took over as chairman.

The FTC took the side of its new head of consumer protection, Timothy J. Muris, who argued against setting rules on food advertising because he said there was little evidence of widespread deception. He said the FTC instead should pursue individual cases vigorously where consumers are being fooled by untrue advertising.

That view was opposed by FTC staff members Melvin H. Orlans and Robert Cheek, who called Muris' requirement of "hard or convincing evidence" before approving a rule "unduly restrictive, inflexible and legally inappropriate." "Indeed, we question whether any rule would satisfy this standard," they wrote in a memo to the commission.

If passed, the rule would have restricted the word "natural" in ads to foods that neither are processed nor contain artificial ingredients and would have forced the listing of calories content on foods advertised as "low calorie." It also would have required ads that talk about low fat content of foods to include specific data on the amount of cholesterol, saturated fat, polyunsaturated fat and total fat content.

Commissioner Michael Pertschuk, FTC chairman in the Carter administration, called yesterday's vote an "abdication [which] will invite a free-for-all for deceptive health claims for food--claims which will cynically exploit and distort growing public concern with diet and health."

As part of the debate over the food-rule proposal, Muris set out for the first time standards that he thinks should be applied before the FTC issues a broad set of regulations covering a large area such as health-food advertising.

Under the Muris standard, proposed new rules should be written only if there is proof that the proposed rule is needed and that it will ease the problem in a way that the benefits exceed the cost. "Only solid statistical evidence can normally provide us information about the magnitude and scope of a particular problem," said Muris, who argued against the use of anecdotal information from individual consumers