Objective To examine the potential of anthropometry as a tool to measure gender discrimination, with particular attention to the WHO growth standards.

Design Surveillance data collected from 1990 to 1999 were analysed. Height-for-age Z-scores were calculated using three norms: the WHO standards, the 1978 National Center for Health Statistics (NCHS) reference and the 1990 British growth reference (UK90).

Setting Bangladesh.

Subjects Boys and girls aged 6–59 months (n 504 358).

Results The three sets of growth curves provided conflicting pictures of the relative growth of girls and boys by age and over time. Conclusions on sex differences in growth depended also on the method used to analyse the curves, be it according to the shape or the relative position of the sex-specific curves. The shapes of the WHO-generated curves uniquely implied that Bangladeshi girls faltered faster or caught up slower than boys throughout their pre-school years, a finding consistent with the literature. In contrast, analysis of the relative position of the curves suggested that girls had higher WHO Z-scores than boys below 24 months of age.

Conclusion Further research is needed to help establish whether and how the WHO international standards can measure gender discrimination in practice, which continues to be a serious problem in many parts of the world