To compare estimates of under-nutrition among pre-school Pakistani children using the WHO growth standard and the National Center for Health Statistics (NCHS) reference.

Prevalence of stunting, wasting and underweight as defined by WHO and NCHS standards are calculated and compared.

The data are from two cross-sectional surveys conducted in the early 1990s, the time frame for setting the baseline for the Millennium Development Goals: (i) National Health Survey of Pakistan (NHSP) assessed the health status of a nationally representative sample and (ii) Thatta Health System Research Project (THSRP) was a survey in Thatta, a rural district of Sindh Province.

In all, 1533 and 1051 children aged 0–35 months from national and Thatta surveys, respectively.

WHO standard gave a significantly higher prevalence of stunting for both national [36·7 (95 % CI 33·2, 40·2)] and Thatta surveys [52·9 (95 % CI 48·9, 56·9)] compared to the NCHS reference [national: 29·1 (95 % CI 25·9, 32·2) and Thatta: 44·8 (95 % CI 41·1, 48·5), respectively]. It also gave significantly higher prevalence of wasting for the Thatta survey [22·9 (95 % CI 20·3, 25·5)] compared to the NCHS reference [15·7 (95 % CI 13·5, 17·8)]. Differences due to choice of standard were pronounced during infancy and for severely wasted and severely stunted children.

Pakistan should switch to the robustly constructed and up-to-date WHO growth standard for assessing under-nutrition. New growth charts should be introduced along with training of health workers. This has implications for nutritional intervention programmes, for resetting the country’s targets for Millennium Development Goal 1 and for monitoring nutritional trends.