(CNN) - As the presidential campaigns focus their efforts on women as a key voting bloc, a new poll shows women who are registered to vote in a dozen key states have very different priorities than their male counterparts.
The USA Today/Gallup survey found that economic issues are important to both genders, but women see one issue as more important.
- Follow the Ticker on Twitter: @PoliticalTicker
- Check out the CNN Electoral Map and Calculator and game out your own strategy for November.
Nearly four in ten - 39% - of women said abortion is the most important issue in this election, the survey said, while a similar number - 38% - of men ranked jobs as their top issue.
The survey includes voters from the nine battleground states on the CNN Electoral Map, plus an additional three - Michigan, New Mexico, and Pennsylvania - which are close, and considered by the Gallup organization as swing states.
Jobs was the second most important issue to women, scoring at 19%,
Women rated healthcare as more important than men did, 18% to 10%.
Behind healthcare was the economy and equal rights/pay/opportunity, rounding out the top five issues for women. Few women voters rated the other items on the survey - government regulation, education, Medicare, Social Security, and taxes - as their no. 1 issue.
Men, however, rated economic issues as significantly more important. Close behind jobs as a top issue was the economy (at 37%) as well as the federal budget and deficit (at 10%).
Defense, national security, foreign policy, and international affairs also made the men's list but did not rank on the women's list.
The swing state survey found women preferred President Barack Obama to GOP nominee Mitt Romney on birth control policy, unemployment, healthcare, and international issues, while they were split on who would best handle the federal budget deficit. Men also preferred Obama on birth control policy, but preferred Romney on unemployment and the federal budget deficit. Men were split on healthcare and international issues.
The survey was conducted between October 5 and 11 and had a sampling error of plus or minus 4 points. It included 1,023 registered voters in 12 key states reached by telephone.