"What you have to ask yourself is who you believe would be the better nominee to go toe-to-toe against John McCain," Clinton said after beating Obama in Pennsylvania last week.
Clinton can't win the most delegates in the primaries, so she is framing the choice for the party's superdelegates - who will likely settle the nomination - around that question.
Obama, who would become the nominee if the superdelegates simply ratify the primary and caucus results, is playing along, though his performance over the last few weeks has done little, if anything, to advance the case for his electability.
Deciding which Democrat has the better chance of winning the presidency isn't nearly as simple as it might seem, according to independent analysts and party strategists who aren't working for either candidate.
"It's more or less pick 'em, if you were a handicapper," said Andrew Kohut, whose independent Pew Research Center has polled nationally and in battleground states. "You'd say you can't make the case for either one."
National opinion surveys illustrate his point. In the Gallup Poll's latest election match-ups, Clinton and Obama are both locked in statistical ties with McCain ("That's much closer than you'd expect for a Republican candidate, given all the Democratic advantages" this election year, Kohut pointed out).
At the moment, concluding which candidate is more electable appears more of an art (that is, educated guesswork) than science.
A highly unscientific sample of those interviewed for this article came down on the side of Obama's superior electability, but the view wasn't unanimous. Some saw him as only marginally stronger, and no one said that Clinton's argument was unreasonable.
"It's not crazy. It's not an absurd case. It's a debatable question," said Stuart Rothenberg, who publishes an independent election newsletter. "There are clearly some Democrats in some key states who aren't going to vote for Obama: older voters, blue-collar voters, the old Reagan Democrats."
Clinton's strategy is to keep the race going as long as possible, in hopes that doubts about Obama will grow. Already, his position "has been eroded" by losing big industrial state primaries, said Rothenberg. It "reminds people who weren't paying attention that he has trouble with a key Democratic voter constituency" - working-class voters.
Alan Secrest, a Democratic pollster, said it was "absolutely unknowable" at this point who would be a stronger nominee but that the picture might be clearer by the time the primaries end in early June.
Obama "sometimes does better with independents and some Republicans," said Secrest. But once the Republicans "have worked him over, I suspect there'd be relatively little difference" between him and Clinton.
What is evident already, said analysts, is that the shape of the fall campaign will depend on which Democrat heads the ticket.
An Obama-McCain contest would be unlike any in recent memory and hinge on an unusually large group of independent, swing voters. A match between Clinton and McCain would more closely resemble the polarized 2000 and 2004 elections, a battle to turn out each party's voter base.
Clinton maintains that she can keep more Democrats from crossing over to the Republican side than Obama. But if she's the nominee, "young people will not turn out" and the African-American vote will be depressed, predicted Curtis Gans, director of American University's Center for the Study of the American Electorate.
If Obama is the candidate, "there will be some people who will not vote for him because of race," he said. "But most of the Democrats will come into the fold, because of things like the war and the economy."
Last week, each candidate made new pitches to the superdelegates, promoting what they see as their unique advantage on the electoral map, based on states they've won so far. But both camps are stretching the predictive value of primaries, strategists said.
A number of the states won by Clinton, such as New York and California, are reliably Democratic in presidential elections and will be again this fall, regardless of who heads the ticket, they said. Meanwhile, Florida, where she prevailed in a primary that the Democratic National Committee invalidated in advance, will likely go Republican in any event.
Obama argues that he can put "red states" in play that Clinton cannot. He's touting his strength in places, such as North Carolina and Texas, that neither he nor Clinton would have a realistic chance of carrying, the analysts said, unless the election became a Democratic rout.
What has caught the attention of Democratic strategists is a handful of traditional battleground states-including Pennsylvania, Ohio and Michigan-that could well be pivotal and where doubts about Obama's strength have grown in recent weeks.
In Michigan, said longtime pollster Ed Sarpolus, "Barack Obama has to understand that if he is to have any chance of overcoming the racial factor, he has to develop a middle-class message, and he has yet to find that."
In Ohio, Clinton would probably do better than Obama against McCain, said Secrest, who advises congressional candidates in the state. He cautioned that it was "a tough call at this point" but "conceivably" Clinton could carry Ohio and Obama could lose it.
The ultimate question is how much difference the electability argument will make with the roughly 230 uncommitted superdelegates - governors, congressmen, senators and DNC members -with the power to choose any candidate they like.
Many dread their role as deciders, afraid of alienating a vast constituency no matter which way they go, and they aren't engaged at all in the debate the candidates are staging for their benefit, according to Bill Carrick, a Democratic consultant.
These superdelegates, he explained, "are basically younger members of Congress and DNC members." They are so uncomfortable making a decision, he added, that they would prefer making none at all.