Evidence on whether the amount of time children spend online affects their mental health is mixed. There may be both benefits and risks. Yet, almost all published research on this topic is from high-income countries. This paper presents new findings across four countries of varying wealth. We analyse data gathered through the Global Kids Online project from nationally representative samples of Internet-using children aged 9 to 17years in Bulgaria (n=1,000), Chile (n=1,000), Ghana (n=2,060) and the Philippines (n=1,873). Data was gathered on Internet usage on week and weekend days. Measures of absolute (comparable across countries) and relative (compared to other children within countries) time use were constructed. Mental health was measured by Cantril's ladder (life satisfaction). The analysis also considers the relative explanatory power on variations in mental health of children's relationships with family and friends. Analysis controlled for age, gender and family socioeconomic status. In Bulgaria and Chile, higher-frequency Internet use is weakly associated with lower life satisfaction. In Ghana and the Philippines, no such pattern was observed. There was no evidence that the relationship between frequency of Internet use and life satisfaction differed by gender. In all four countries, the quality of children's close relationships showed a much stronger relationship with their life satisfaction than did time spent on the Internet. Time spent on the Internet does not appear to be strongly linked to children's life satisfaction, and results from one country should not be assumed to transfer to another. Improving the quality of children's close relationships offers a more fruitful area for intervention than restricting their time online. Future research could consider a wider range of countries and links between the nature, rather than quantity, of Internet usage and mental health.