This paper explores the effect of work on children's school achievement as measured by the Peabody Picture Vocabulary Test (PPVT). Identifying the causal effects of child work on education is made difficult because the choice of work and/or school is made simultaneously and may be determined by the same potentially unobserved factors.Therefore, both ordinary least square and instrumental variable estimation methods were used to identify the effect of work on school achievement. We used dummy variables for drought, crop failure and pests and diseases, for increases in the prices of food, and for urban locality as instruments which are highly, though not directly, correlated with achievement in education. The results obtained showed that children's work had a negative effect on achievement in education as measured by the raw PPVT score. Therefore, it is important to design mechanisms that enable households to withstand income shocks without resorting to child work. The Government of Ethiopia should consider implementing a programme that provides financial incentives to households to send their children to school regularly, thus potentially increasing the children’s future earning capacity. A conditional cash transfer programme could be a way of helping children achieve better in school and of minimising child work.