Children who play video games for more than three hours a day are more likely to be hyperactive and unruly at school, a study has shown.
The problem behaviour, which included getting into fights and lack of interest in class, is linked to the amount of time devoted to games rather than the types of games played, said researchers.
No clear association was seen between violent games and the extent to which the 12 to 13-year-olds displayed a tendency to fight or show aggression. Some kinds of games, as well as low play levels of less than an hour a day, appeared to have benefits.
Lead author Andy Przybylski, from Oxford University, said: "We can see links between some types of games and children's behaviour, as well as time spent playing.
"However, we cannot say that game play causes good or bad behaviour. We also know that the risks attached to game-playing are small. A range of other factors in a child's life will influence their behaviour more as this research suggests that playing electronic games may be a statistically significant but minor factor in how children progress academically or in their emotional well-being."
Although some parents might believe that playing strategy and puzzle games can boost school grades or increase social skills, this was not borne out by the research.
Children who played such games were no more sociable or better in class than those who did not.
However, certain games were associated with some types of positive behaviour.
Children who played video games with a co-operative and competitive element had fewer emotional difficulties or problems with peers, while th ose who preferred solitary games did well academically, displayed fewer emotional problems, and were less likely to get involved in fights.
It was not clear whether the games led to the behaviour, or whether children of a certain disposition chose particular games.
The research, published in the journal Psychology of Popular Media Culture, was based on teachers' assessments of behaviour at a school in south-east England.
Rather than relying on the reports of children themselves, the teachers were asked to rate the academic performance and conduct of the 200 pupils in the study group.
Children were given numbers so that their identities were not revealed to the researchers. The assessments were matched to the results of a survey of game-playing behaviour conducted among the children.
Co-author Allison Mishkin, also from Oxford University, said: "These results highlight that playing video games may just be another style of play that children engage with in the digital age, with the benefits felt from the act of playing rather than the medium itself being the significant factor."