Bullying is a major public health problem. We aimed to estimate the prevalence of cyberbullying and traditional bullying among adolescents in England, and assess its relative effects on mental well-being. In this population-based study, we analysed data from a nationally representative cross-sectional study, What About Youth, which enrolled a random sample of 298 080 school pupils drawn from 564 886 National Pupil Database records of adolescents aged 15 years, living in England, with matching postcode and local authority data, to complete self-report surveys between Sept 22, 2014, and Jan 9, 2015. Mental well-being, defined as life satisfaction, fulfilling social relationships, purpose in life, and a subjective sense of flourishing, was assessed using the Warwick-Edinburgh Mental Well-being Scale and was compared between those adolescents who reported traditional bullying (including physical, verbal, and relational bullying) or cyberbullying 2-3 times a month or more compared with those adolescents who reported traditional bullying and cyberbullying once or twice in the past couple of months or less. Traditional bullying was defined as repeated, intentional aggression that is targeted at a person who cannot easily defend himself or herself; cyberbullying was additionally defined as taking place in an electronic context (eg, e-mail, blogs, instant messages, text messages). 120 115 eligible adolescents completed questionnaires, of whom 110 788 adolescents completed measures of bullying. 33 363 (adjusted: 30% total, 20 668 girls [36%], 12 695 [24%] boys) reported any form of regular bullying in the past couple of months. 29 302 (27% total, 17 745 [31%] girls, 11 557 [24%] boys) reported physical, verbal, and relational (ie, traditional) bullying only, while 406 (<1% total, 276 [<1%] girls, 130 [<1%] boys,) reported only cyberbullying, and 3655 (3% total, 2647 [5%] girls, 1008 [2%] boys) reported both traditional and cyberbullying. Both kinds of victimisation were related to poorer mental well-being (adjusted analyses, traditional: b coefficient=-1·99 (SE 0·001); cyberbullying: b coefficient=-0·86 (0·06). Cybervictimisation accounted for less than 0·1% of observed variability in mental well-being compared with 5·0% of variability accounted for by traditional victimisation. Traditional bullying is considerably more common among adolescents in England than cyberbullying. While both forms of bullying were associated with poorer mental well-being, cyberbullying accounted for a very small share of variance after adjustment for offline bullying and other covariates. None.