It’s an oft-cited paradox that wildland firefighters light fire to fight fire. If it’s black, it’s not as likely to burn, so backburns, burnouts and backfires can be effective tools to widen fire lines, starve massive blazes of fuel, even save lives.

In an emergency, lighting a backfire makes sense. Viewers of 1960s television may remember Lassie heroically lighting a backfire to save an injured U.S. Forest Service ranger in one episode. The flames roared into the head of the main fire, consuming the fuel that would have allowed the forest fire to burn unchecked. Or there’s the case of Wag Dodge, one of three survivors of the Mann Gulch Fire in 1949, who lit a small fire in front of the blowup bearing down on the crew, hunkered in the ashes of the small fire and rode out the passing firestorm.

Recently, though, backfires have made news of another kind. When a prescribed fire burned out of control in Los Alamos, New Mexico, in 2000, it was a backfire set to fight the prescribed fire that burned into the town, destroying hundreds of homes and damaging the national laboratory. In Montana, after dramatic fires later that same year, a group of homeowners sued the Forest Service for $54 million to compensate for damage to their homes caused by a backfire. Jerry and Gayle Sorensen (Who Needs Help Like This?) describe another frightening episode with a backfire during southern Oregon’s Biscuit Fire of 2002. For the Sorensens, Forest Service firefighting left them wondering what the agency was trying to accomplish by setting a backfire a few hundred yards from their house in the middle of a hot, dry afternoon.

Lighting, and studying, backfires is an imprecise science. “You never know where the fire is going to stop,” says Tom Halferty, a biologist for the National Marine Fisheries Service who studies the ecological effects of fire, and firefighting, on fish. “It’s a judgment call.”

Backburns are intended to starve a bigger fire of fuel. Firefighters try to light backburns that consume most of the fuel in a designated area, and one could assume that a backburn that leaves any significant fuel behind would be an unsuccessful backfire, says Tom Atzet, a Forest Service ecologist in southwestern Oregon. To that end, backburns are often set when conditions are ripe for consuming fuel. Fires burning naturally may or may not consume fuels in a landscape as thoroughly, says Atzet, and studies trying to confirm whether they do or not have been inconclusive.

Halferty is putting together a report that describes the ecological effects of fighting the Biscuit Fire. Teasing out where and how fires within and near the main fire started, though, is tricky because where and how much was backburned while crews fought the main fire wasn’t documented systematically. Halferty has had to assemble and compare the stories of a number of on-the-ground firefighters and resource advisers, and from that, he’s putting together a bigger story of how the fire was fought.

Halferty won’t comment about the extent of what in the perimeter of the nearly 500,000-acre Biscuit Fire was backburned and what was the main fire until he releases the report, but his work illustrates a key point. Once a fire is going, there are few systems in place to chart how the resource is being managed while it’s burning and what the consequences of that management might be. The teams fighting the fire don’t do long-term resource work; they dig line, backburn, bulldoze and drop retardant-and treat every fire as an emergency.