They died on the battlefield at the hands of their countrymen or they succumbed to disease and untreated wounds -- at a rate of several hundred men a day.

When the Civil War ended in 1865, no one was sure exactly how many lives had been lost as a result. Official estimates put the death toll at 620,000, where it has stood for more than a century. Many historians, scholars and even the men who came up with the figure always assumed the actual figure was much higher, but they had no basis to revise it.

Until now.

Binghamton University historian David Hacker has used 19th century census data to calculate a new estimate of the Civil War dead, and it is much higher than anyone thought. Hacker said his research suggests 752,000 men died as a result of the war, but the number could be as high as 851,000.

Hacker, who will publish his findings in the December issue of Civil War History, said he originally started out researching 19th century mortality rates.

"The traditional estimate has become iconic," Hacker said. "I was originally researching census under counts. If you know mortality rates, it's possible to get a better estimate. I was going with the 620,000 and what I was coming up with didn't make much sense."

Hacker used census microdata that had been compiled by the University of Minnesota to look at the decades before and after the war to establish normal survival rates for men and women. Then he used that pattern on the 1860s.

Based on that, Hacker said, it appears that 650,000 to 850,000 additional men died in that decade, giving him an average of 750,000.

Because it is math, not actual tracking of individuals, there is no way to put a number on which side these additional dead come from, but it's clear to some historians that the Confederate estimate is very low. Many of the South's records were destroyed, and U.S. Army officials claimed Confederate Gen. Robert E. Lee often purposely under-counted casualties in battle.

Original estimates were lower than the widely accepted 620,000 number, but were revised upward as pension claims began to trickle in over the years.

At Fort Sumter, historian Richard Hatcher said he expects this report will spark discussions in Civil War circles, particularly because the number has always been such an interesting topic for debate.

"Over the last 20 years, most people have settled on 600,000 to 620,000 or 625,000 -- even 650,000," Hatcher said. "I thought it was a good working number. But I would not be surprised if some academic study said it was higher."

Hatcher said tracking people, even soldiers, in the mid-19th century was difficult. Some men just disappeared. They could have been killed by the enemy and buried anonymously, ending up listed as missing in action as opposed to a casualty. And, he noted, it was easy to get lost in those days.

Of the 620,000 number, less than half -- 260,000 -- are Confederate casualties. Hacker said because of his methodology and available data there was no way to break down the number of deaths by state or even between Union and Confederate. James McPherson, the Pulitzer Prize-winning Civil War historian, said that he believes the difference in the accepted number and Hacker's findings is probably related to the under-reporting of Southern deaths in the conflict.

"Even if it might not be quite as high as 750,000, I have always been convinced that the consensus figure of 620,000 is too low, and especially that the figure of 260,000 Confederate dead is definitely too low," McPherson said.

Hacker said the inexact and lost records mean that there will never be a completely accurate picture of just how many lives the war cost.

"There's always soldiers you're going to miss," he said. "You're always going to be low."

Reach Brian Hicks at 937-5561 or follow him on Twitter at @BriHicks_PandC.