The federal government has granted its first approval of a private company's plan to grow outdoors plants that have been produced through gene-splicing techniques.

The National Institutes of Health said yesterday it has a approved a proposal by Agracetus, a joint venture between Cetus Corp. and W. R. Grace & Co., to test tobacco plants grown from seeds that have been genetically engineered to make the plants disease resistant.

The decision is encouraging for biotechnology companies eager to apply gene-splicing techniques to agriculture, but comes at a time of renewed public debate about the government's abilities to evaluate the environmental risks of releasing genetically altered organisms outside laboratories.

Two congressional panels have scheduled hearings on the subject, and the U.S. Environmental Protection Agency is expected today to grant its first approval of an outdoor test of live bacteria that have been genetically altered to act as a pesticide.

NIH said yesterday that the Agracetus experiment would "present no significant risk to health or the environment."

Jeremy Rifkin, whose legal challenge of other genetic engineering experiments has delayed approval of the Agracetus plan, said "We don't believe any agency should go ahead and release any genetically modified organism -- microbe, plant or animal -- into the environment until they have developed the beginnings of predictive ecology methods, a science to judge the risks."

Agracetus said yesterday that it has not decided whether it will go ahead with the experiment. If it does, it will wait until spring to plant tobacco on a 70-foot by 80-foot plot in Wisconsin. The company would grow the plants from seeds that have genetic material inserted in order to increase the plants' resistance to a disease called crown gall. The company will harvest all but five plants before they flower and will bag the remaining five prior to flowering to prevent pollen dispersal.

Every year, agricultural companies test new types of plants developed through traditional cross-breeding techniques to grow bigger, stronger or more productive.

Many of these companies view recombinant DNA techniques as a way to accomplish the same goal more easily, by directly manipulating the genes involved with specific desired traits. These agricultural biotechnology companies are eager to use gene-splicing to produce crops that need less fertilizer or fewer nutrients, or plants that produce substances that kill insects or resist diseases.

The NIH decision reflects both the progress and the continuing murkiness of the Reagan administration's system for regulating the commercial products of recombinant DNA techniques -- the manipulation of the genetic code carried by DNA, or deoxyribonucleic acid.

No law requires a private company to seek federal permission to conduct an outdoor test of plants created through gene-splicing. But Agracetus, based in Middleton, Wis., voluntarily asked NIH to review the experiment in May 1983.

Winston J. Brill, Agracetus vice president and director of research, said the company turned to NIH for two reasons: it wanted to allay public concerns about a new technology by submitting its test for public review by a competent federal agency, and NIH was at that time the only federal agency with expertise and experience in evaluating recombinant DNA research.

Today, the U.S. Department of Agriculture has assumed responsibility for reviews of recombinant DNA-produced plants, but does not require companies to submit their plans for approval. By contrast, EPA requires notice of any plans to release genetically engineered microbial pesticides, and the Food and Drug Administration must approve any tests of genetically engineered medicines or food additives.

Agracetus's proposal had to wait 2 1/2 years while NIH fought a lawsuit challenging its methods of approving another experiment that would have released into the environment a genetically altered living microbe used as a pesticide.