Many families employ private tutors to help children with their schoolwork, thus participating in a 'shadow education' system that supplements normal schooling. International surveys show that there is wide variation in the extent of tutoring in different countries. This paper considers reasons for this variation and evaluates evidence on the prevalence and effectiveness of private tutoring. The effectiveness of tutoring is mixed, with some well-designed programmes achieving large gains in attainment, whereas surveys show little impact. Some of these discrepancies arise from weak conceptualisation and methodological issues. It is argued that quality indicators should be added to analyses of survey data to provide more reliable estimates. Evidence from research on extra-curricular activities is used to explore the 'penumbra' between public and private sector involvement. Systematic monitoring of the shadow system is recommended. Policies should be developed, as it is likely that the prevalence of private tutoring will increase.