The purpose of the current study was to examine the frequency of cyber bullying among youth by distinguishing among the three categories of involvement in cyber bullying: victims, bullies, and bully–victims, to compare these to a fourth category of students who are not involved in the three categories of cyber bullying and to explore the factors that contribute to involvement in cyber bullying. This study utilized a large and diverse sample of 2186 middle and high school students, who completed self report questionnaires during class time. We performed a Multinomial Logistic Regression to examine the relationship between the cyber bullying categories and our independent variables (gender, age, technology use, parental involvement and safety). Over 30% of the students in this study identified as involved in cyber bullying, as victims or perpetrators, and one in four of the students (25.7%) reported having been involved in cyber bullying as both bully and victim during the previous three months. Students who were involved in cyber bullying were more likely than others to report perpetration of violence toward peers, to use computers for more hours a day, and to give their password to friends. Other risk factors, such as gender, age and safety, were found to be specific only for one category of cyber bullying. The findings revealed that students are highly involved in cyber bullying. Several unique characteristics emerged regarding the frequency and risk factors of students' involvement in cyber bullying. In traditional bullying the category of bully–victims represents the smallest and most vulnerable group of children, whereas in the current study the bully–victims category emerged as common. In addition, females were more likely than males to be bully–victims, in contrast to research on traditional bullying, in which more males than females are typically involved as bully–victims. In addition, several risk factors were common among the three groups of children, including the amount of hours per day students use the computer, and giving passwords to a friend. These results point to the need for further examination and to focus on the risk factors for students' cyber bullying involvement in each of the three categories. ► We compare cyber bullying: victims, bullies, bully–victims, no involvement. ► We explore the factors that contribute to involvement in cyber bullying. ► One quarter reported involvement as both bully and victim in previous three months. ► In traditional bullying bully–victims represent the smallest most vulnerable group. ► In the current study the bully–victims category emerged as more common.