Inclusion of sporting influences into a measure of defensive weapon ownership and errors in attribution of demographic variables have weakened the results of many previous studies of defensive weapon ownership. In order to remedy these flaws, this study used only those respondents who claimed to live alone, and respondents were specifically asked if they had purchased a weapon for self-protection. Data for this analysis were taken from a subsample of that used in Smith and Uchida's (1988) analysis of defensive weapon ownership. The Police Services Study was completed in 1977 and consisted of interviews with a random sample of residents from the Standard Metropolitan Statistical Areas of Tampa-St. Petersburg, Fla.; Rochester, N.Y.; and St. Louis, Mo. There were 2,386 individuals identified as living in single adult households at the time of the survey; these were selected for inclusion in the analysis. The independent variables measured were gender, age, race, income, education, home ownership, communication among neighbors, rating of police, previous victimization, high perceived crime risk, perception that crime is increasing, and any children in the household. Separate probit equations were estimated for males and females to assess differential indicators for men and women. Results show that although having a prior victimization was a significant positive indicator of self-help for men, women did not show similar relationships. Variables that deal with neighborhood characteristics, such as the percentage of single-parent households with children aged 12 through 20 and perceptions that crime is increasing in the neighborhood were significant predictors of defensive weapon ownership for women. 3 tables, 6 notes, and 18 references