I've managed to crunch the numbers from both Male and Female National Survey of Family Growth. As a treat for sdaedalus, I've included the male data.
First My approach to analysis was Catholic, in that you're only allowed to get married once. Remarriages count as a fail. However the data from NSFG 2002 included remarriages amongst the currently married.
Amongst males, 21% of the currently married were second or later marriages.
Amongst females, 23% of the currently married were second or later marriages.
What I basically attempted to do was calculate the following:
%Married=%married/(%married+%divorced+%remarried) for each sexual partner cohort.
1) The first thing to work out was how many remarriages there were.
2) Then I proportionately distributed the remarriages amongst the the greater-than-two-sexual-partner cohorts.
3) Then I subtracted the remarriages from the current marriages in these cohorts to give me a estimate of married once in each group.
4) Then I added the remarriages to the divorce group and performed the above calculation.
Note, this graph does not measure how long the the subjects had been married, simply their marriage status by partner count.
It's interesting to note that male promiscuity does not seem to affect marital stability as much as female.