There is evidence that suggests "no-drop" policies do keep more cases in the system than do noncoercive policies; however, coercive policies may also produce lower rates of successfully completed prosecutions. In San Diego, for example, prosecution data for 1996 show a dropped-charge rate of 33 percent, with another 34 percent of victims either recanting or testifying on behalf of the defense. Only one out of three cases involved "cooperative victims." In Seattle, under a 1996 no-drop policy, 48 percent of the cases filed were dismissed, and 40 percent of those prosecuted had "favorable dispositions." Data were not reported on how many of the victims in favorable-outcome cases cooperated with the prosecutor. A more recent study of no-drop policies in four jurisdictions found that no-drop policies increased convictions. The increase was associated with rigorous screening to determine the likelihood of conviction. As of now, however, few research studies have assessed any aspect of domestic-violence prosecution. In order to know with certainty that prosecution actually protects victims, research must compare cases randomly assigned to either prosecution or non-prosecution, with controls for court outcomes. To date, no such research exists. There is an urgent need for research that tests the protective impacts of not only coercive prosecution policies by especially policies that give top priority to the interests of battered women. 34 references