Dan Olweus' pioneering work on traditional bullying laid the foundation for how cyberbullying is defined and measured today. Many researchers believe that these two forms of bullying share some similarities. However, two perspectives have been outlined for the relationship between traditional bullying and cyberbullying--the extension perspective (predictors and outcomes of cyber and traditional bullying are similar) vs. the differences perspective (cyber and traditional bullying have unique patterns of relationships with predictors and outcomes). In the current study, we attempted to shed light on the debate about these two perspectives by addressing three questions: (1) Do cyber and traditional bullying have similar or unique predictors? (2) How do cyberbullying and traditional bullying relate to outcomes? and (3) Is there a bidirectional relationship between cyber and traditional bullying and "Outcomes"? We collected online survey data across two waves from 151 college students at two universities located in the US (n = 118 females, n = 32 males, n = 1 preferred not to disclose, M age = 21.68 years old, SD age = 3.73 years), focusing on known predictors, cyber and traditional forms of bullying, and outcomes. Results indicated relatively more support for the extension perspective across our three research questions: (1) Relatively more variables predicted both cyber and traditional bullying (self-esteem, school demands, university rule clarity, Machiavellianism, and psychopathy); (2) Both cyber and traditional bullying predicted similar outcomes (including deviant behavior, helping behavior, symptoms of anxiety, and symptoms of depression); and (3) Helping behavior and number of alcoholic beverages consumed predicted both cyberbullying victimization and traditional bullying victimization. Findings are discussed in terms of their implications for future research and practice.