Examination of the longitudinal relationship between empathy, social-emotional problems and cyberbullying is still rare and the present study is one of very few. The present study assessed whether low scores of affective and cognitive empathy at wave 1 (t1) can predict involvement in cyberbullying five months later (t2). Furthermore, it was examined whether involvement in cyberbullying at t1 predicts psychopathological symptoms and social withdrawal at t2. Participants were 77 7th and 8th grade students from a control group of a pre-/posttest short-term longitudinal evaluation study of a general anti-bullying program (mean age[subscript t1] = 12.53 years, SD = 0.68; gender[subscript t1] = 54.5% boys, 45.5% girls). Separate quasi-poisson regression analyses were conducted and traditional bullying and victimization were included as control variables. Low scores of affective, but not cognitive, empathy predicted cyberbullying but not cybervictimization at t2. Neither cyberbullying nor cybervictimization predicted social withdrawal or psychopathological symptoms at t2 as assessed in this study. The research hypotheses were only partly supported, however, this study using short-term longitudinal data revealed evidence for the importance of (affective) empathy in cyberbullying perpetration.