Drawing on prior research on disputes, the author hypothesizes that this effect depends on the way a claim is framed (which in turn is affected by the claimant's gender), on the mode of attorney involvement, and on claimant use of informational networks. These hypotheses are examined through an analysis of interviews of 96 recently divorced men and women in Cook County, Ill. The dependent variable in the model is whether or not law was perceived by the respondent as being used in the molding of the final divorce settlement. Independent variables pertained to the characteristics of the negotiations that led to custody and child-support arrangements. The hypothesis that women are more likely to frame their divorce issues relationally than men is not supported by the data. The hypothesis that those who initially choose a legalistic framework would be more likely to use the law than those initially choosing a relational framework was not supported by the data. One third of both those who initially chose a legalistic framework and a relational framework ultimately reported using the law in their negotiations; two-thirds of both groups reported not using the law. Those who chose an initial legalistic framework more often allowed their lawyers more scope and used them as negotiators than did the relational framers. Among those who were initially legalists, getting information from personal networks made little difference in the proportions who later used the law in their negotiations. For those initially using a relational framework, those who got information from their personal network were almost twice as likely to use the law than those who did not get information from their personal network. 3 tables and 30 references