Research that is soon to be published shows no significant improvement in the volume of oxygen-depleted water in the Chesapeake Bay since the inception of the bay's cleanup in the mid-1980s, a finding that runs contrary to reports from the government agency leading the estuary's restoration.

The research on oxygen-depleted water, considered a leading indicator of bay health, again highlights what many environmentalists describe as the Chesapeake Bay Program's unjustifiably positive reports about the cleanup's progress.

The Bay Program's Web site on oxygen-depleted water says "there are recent indications of an improving trend since 1985."

But a paper by two University of Maryland scientists reviewing similar data sees no trend at all in the key indicator since 1985.

"The take-home message is, since 1985, it's been pretty bad," said Walter R. Boynton, a scientist with the university's Center for Environmental Science and coauthor of the paper with James D. Hagy. The study will be in the August issue of the journal Estuaries.

Exactly how much the Chesapeake Bay cleanup has progressed since its inception in the mid-1980s is disputed, and this difference over a key statistic -- measuring oxygen-depleted water -- promises to open a new front in the debate.

Environmentalists have long noted that while the Bay Program's computer estimates reported significant improvements in pollution flows to the estuary, the bay suffered precipitous drops in its oyster and crab harvests and, at times, its desolate dead zone -- where there is too little oxygen for most bay life -- expanded.

The Bay Program's computer estimates of pollution have significantly understated pollution flows, program scientists have conceded, leading to revisions of progress reports twice in recent years.

Now the recent university research on oxygen-depleted water, running contrary to the Chesapeake Bay Program's findings of progress, again focuses attention on what some describe as the Bay Program's overly positive reviews.

"Unfortunately, this fits with a pattern we've been seeing," said Theresa Pierno, a vice president of the Chesapeake Bay Foundation, an educational and watchdog group. The Bay Program "continues to issue optimistic reports showing improvement when in fact the data we and others have seen is not showing that. People need to know we are not winning this battle," she said.

A spokesman for the Chesapeake Bay Program, Christopher S. Conner, said the findings of improvement are based on sound science and that "the most important message we have is that conditions in the Chesapeake Bay have to improve."

Since the 1980s, Virginia, Maryland, Pennsylvania and the District have sought to reduce the amounts of phosphorus and nitrogen from farms, treatment plants and other sources that eventually enter the bay.

Nitrogen and phosphorus pollution sets off algae blooms that draw oxygen out of the water, starving sea grass, crabs and other creatures of the essential element. Reducing the flows of nitrogen and phosphorus to the bay, in turn lowering the volume of oxygen-depleted water, has been the central focus of the bay restoration.

According to the Bay Program's computer estimates, which have been published in its State of the Chesapeake Bay reports and touted at news conferences, the flow of phosphorus has dropped 28 percent since 1985 and the flow of nitrogen has declined 18 percent.

The statistics on oxygen-depleted water are considered another key measure of the bay cleanup's progress.

The Bay Program says that the volume of oxygen-depleted, or hypoxic, water has shown a downward trend since 1985. The University of Maryland researchers found no trend since 1985. Under a longer view, the data showed an increase in the problem that has accelerated in recent years.

The discrepancy in the findings appears to arise from a difference in how researchers defined "oxygen-depleted." Scientists agree that when measuring from 1985 there has been no sign of improvement in hypoxic waters in the bay under the usual definitions.

The Bay Program's report of improvement relies on a broader definition of hypoxic, which some scientists believe is a valuable gauge.

"We used different statistical methods and we looked at different levels of oxygen," said David Jasinski, a water monitoring data analyst with the Environmental Protection Agency's Chesapeake Bay Program office. "There is an improving trend."

He described the trend as "slight."

The divide over the amount of oxygen-depleted water echoes the debate over whether the primary measure of bay progress should be the Bay Program's computer model or actual water samples.

While the computer model shows significantly reduced pollution flows, observed water monitoring data from rivers entering the bay indicate far less progress, particularly for phosphorus.

Program officials note that the river sampling data from the U.S. Geological Survey cover about 80 percent of the watershed and miss significant pollution reductions at many improved sewage treatment plants. The model, they have said, takes the entire bay into account and can adjust for rainfall fluctuations, which affect pollution flows.

But the model has proven an inconsistent guide.

Donald F. Boesch, president of the University of Maryland's Center for Environmental Science, has criticized the Bay Program for what he calls "the over-reliance on the virtual reality of models."

The assumptions behind the computer model likely will be revised again within three years, and again could shrink the amount of progress reported, said Tom Simpson, a consulting scientist to the program from the University of Maryland.

"The more we learn about this, the more we learn our [assumptions about agricultural pollution] have been too optimistic," said Simpson, who is working with the Bay Program to adjust its assumptions about agricultural practices. "I'd expect the revisions will have to reflect that."