What happens when a newspaper's star investigative reporter decides to write a book about the man in charge of the nation's darkest secrets?

When does his research go into the newspaper and when can it be held until publication of the book? When does the obligation to newspaper readers for a daily story outweigh the writer's desire to get more information for a future story?

Ever since Washington Post assistant managing editor Bob Woodward began his research in late 1984 on William J. Casey and his directorship of the Central Intelligence Agency, Woodward and his editors at The Post have tried to divide his findings in such a way that would satisfy the needs of the paper and his publisher.

The result was that Woodward produced 75 stories in The Post since January 1986 while writing his book, "VEIL: The Secret Wars of the CIA, 1981-1987," which arrived in Washington bookstores Monday afternoon.

In the three years he was investigating Casey's tenure at the CIA, Woodward wrote newspaper stories that included scoops on the Reagan administration's disinformation campaign designed to rattle Libyan leader Moammar Gadhafi, new information about Pakistan's nuclear capabilities and detail about the CIA's view of Gadhafi's mental state.

Nevertheless, when the book was excerpted in The Post, Newsweek and other newspapers starting Sunday, there were enough new details to raise the issue among journalists and some politicians of why some items were not published in the paper as Woodward learned them. As New York Times columnist Flora Lewis wrote in yesterday's editions, echoing the key question of Watergate: "What did the editors of The Washington Post know and when did they know it?"

Among the revelations in the book that were not published earlier in the newspaper is a dramatic hospital scene last winter when Casey seemed to acknowledge that he knew about the diversion of funds from U.S. arms sales to Iran to aid the Nicaraguan contras.

Another is the news that the Saudi intelligence service helped Casey with three covert operations, including an effort to assassinate and then bribe Sheik Mohammed Hussein Fadlallah, the Hezbollah leader believed to be behind bombings of American facilities in Lebanon. Woodward had written in May 1985 that an assassination attempt against Fadlallah had been carried out by a team with an indirect connection to the CIA that was on a "runaway mission." The attempt failed and 80 people were killed in the car bombing.

"It isn't enough that The Washington Post, thanks to Bob Woodward, got all these stories first," said Post executive editor Benjamin C. Bradlee. "It's that we didn't get them to fit some schedule that the critics think was more appropriate."

"Our readers and we at The Post have benefited enormously from this project over the life of it," said Robert G. Kaiser, The Post's assistant managing editor for national news. "It's had a huge payoff."

Woodward's view, as explained in several interviews over the last few days, is that a reporter often holds out information in hopes of gleaning more information from a source. His own method is to interview and reinterview his sources, comparing bits from one interview with pieces from another.

"You have to get the kind of evidence that is persuasive as a story," Woodward said. "On Casey {and the Iran-contra diversion}, I don't have something conclusive. I'm still working on it. On the relationship to the Saudi intelligence service, I believe I do.

"The diversion is still a long-term story. It has not been answered, yet," Woodward said. "Some people say that if Casey died, well, then that's the end of it, but maybe he kept a diary, maybe he talked to somebody . . . . The investigation and the inquiry go on."

The Casey hospital scene, which has been denied by his widow, Sophia Casey, and reaffirmed by Woodward, who said she was not there at the time, was not included in The Post's Iran coverage previously, Woodward said, because he felt it was not a clear answer from the former CIA director about his knowledge or involvement.

"It didn't pass the threshold test for a news story," Woodward said yesterday.

Woodward said that in February or March he wrote a draft of the scene and talked to his editor at Simon & Schuster Inc., Alice Mayhew. "She and I agreed it was not conclusive but in the cumulative portrait {Woodward} had built of Casey, it fits," he said.

When the fired National Security Council aide, Lt. Col. Oliver L. North, testified in July that Casey knew about the Iran-contra affair, Woodward wrote that officials who worked with Casey said that North's description "could be true" but added that some also said that Casey, who had died in May, "is a convenient cover and scapegoat for North."

Asked about why he could not use the Casey bedside interview as part of that story or in a "sidebar" or smaller story attached to it, Woodward said that the statement was early in the hearings, and "I expected that they would question him more, that they might have a document.

"As evidenced by that story, I was still talking to people who knew and worked with Casey, and they gave me those observations. To throw in a paragraph that said 'Oh, by the way, I went to his hospital room' -- it just didn't dawn on me."

Woodward said that by then Bradlee and Kaiser had read the book with that last scene in it, and the issue of adding the account was not raised during discussion about the North story.

On Fadlallah, Woodward and Post reporter Charles R. Babcock wrote in May 1985 that an effort to assassinate the Hezbollah leader had been a "runaway mission" of the CIA that had accidentally resulted in the death of 80 people. He said that he did not know of the Saudi intelligence efforts to kill or bribe Fadlallah until last July.

Woodward said that he found the key to Casey's operation was the "black holes," the use of intelligence units from other countries to do "the dirty work." He said he spent months trying to find out if the Israelis were behind the Fadlallah attempt.

"Finally Casey told me it was the Saudis," Woodward said, but he said that was still not enough confirmation for him to do a story.

Woodward said he learned in July about details of the Saudi connection, including a scene involving Casey and the Saudi ambassador to the United States, Prince Bandar bin Sultan, at Bandar's palatial residence. In the garden, "when they were about as far away as possible from the house and the security guards," Casey gave Bandar a small card with a Swiss bank account number. It was the account where $3 million was to go; $2 million later was used to bribe Fadlallah to end the car bombings, the book said.

Woodward's decision to write a book about Casey and the CIA came after he finished a book about actor John Belushi. He and Bradlee talked about his next book project and, Woodward said, Bradlee encouraged him to write about Casey.

"He and I agreed we would be getting the fruits of his labor regularly," Bradlee said. "Woodward and I have been dealing together a long time. These were not hostile negotiations here, but on the question of whether he kept anything for the book, I'm sure he did."

Woodward said that during this period there were two stories that he argued should be held out of the paper and that editors wanted to publish -- one about the 1981 Libyan "hit squads." The other was about a CIA "payoff" of $100,000 to Eugenia Charles, the prime minister of Dominica, who headed a group of Caribbean leaders who requested the U.S. invasion of Grenada.

Woodward said that at the time, he felt he did not have the Charles story "nailed down sufficiently" and that there were strong denials from her and the White House. On the Libyan situation, he said that last year he preferred to report on what was happening at the time concerning Libya, rather than events five years earlier.

Although critics have focused on whether Woodward "hoarded" news for the book, he also occasionally was a prime force in trying to get stories into the paper -- sometimes in opposition to Post editors.

For example, Woodward pushed to get The Post to publish information about Ronald W. Pelton, the former National Security Agency official convicted of espionage. Before and during Pelton's trial last May, The Post finally published stories by Woodward and reporter Patrick E. Tyler, with some technical detail deleted, about an intelligence-collection operation involving U.S. submarines.

"We kept out these technical details because we were not convinced that the whole operation was completely blown by Pelton," said Post managing editor Leonard Downie Jr. "As long as we couldn't be sure, we didn't want to publish {the paragraphs} for fear that we would tell the Russians something they didn't already know. Contrary to popular belief, we do care about national security."

Woodward, who has been at The Post since 1971, has a special niche at the paper as the reporter who helped break many of the major Watergate stories with former Post reporter Carl Bernstein. Kaiser called him "a force of nature" at The Post, where performance has earned its privileges.

"You cannot have the best people doing the same thing for 30 years," said Bradlee. "You've got to create opportunities, whether it's a leave or book or sabbatical -- whatever it is that maximizes chances of keeping these people interested and productive.

"Woodward has a special position at The Washington Post, and it is a tremendous advantage to The Washington Post and to its readers," he said.

The Post's arrangement with Woodward for this book was that he would remain on the staff at full pay and would write or help with stories that surfaced during his research.

Woodward said The Post bought first serial rights to the book for $1 and resold them to other publications, including Newsweek for $75,000 and 46 newspapers in the United States and Canada for varying rates. Woodward declined to say how much Simon & Schuster paid him as an advance for the book, but there are reported estimates that he received more than $1 million.

Charles Hayward, publisher at Simon & Schuster, said yesterday that Woodward's book was published on a "crash cycle" that took about a month, instead of the normal nine months, from the time a finished, edited manuscript can be converted into a hardbound book and distributed to bookstores.

Although Simon & Schuster printed 601,000 copies, Hayward said he is considering ordering a second printing, possibly as early as today. He said that sales could number as high as 1 million and "early indications are what we expected -- that the book would be flying out of the bookstores."