This paper investigates the effect of child labour on child schooling using the 2002 dataset from the
Bangladesh National Child Labour Survey (NCLS). The study provides evidence on the sensitivity of
the results to the methodology adopted and the gender of the child. Working hours have an adverse
effect on the child’s schooling outcome and the result is robust to the procedure used, but the precise
nature of the relationship varies between the parametric and non-parametric approaches. While the
former suggests a U-shaped relationship between working hours and child schooling, there is a
monotonic and inverse relationship between the two in case of the latter. The study also provides
evidence that suggests that the gender of the child affects the nature of the relationship, as do the
parental characteristics, especially parental education. These results attain special policy significance
in view of the evidence that both parents show a significant preference for educating a female child
that appears to be at odds with the picture of gender bias in favour of the male child in South Asia
portrayed in the literature.