AbstractObjective:To assess whether the observed prevalence of the double burden of malnutrition (DBM) would be higher than expected on the basis of chance, through analyses at national, wealth quintile and individual child levels.Design:We selected nationally representative surveys from low- and middle-income countries (LMIC) carried out since 2005 with anthropometric measures on children under 5 years of age. Household wealth was assessed through asset indices. The expected prevalence of DBM was estimated by multiplying the prevalence of stunting (low height/length for age) and overweight (high weight for height/length). The WHO recommended cut-offs (20% for stunting and 10% for overweight) that were used to define DBM at national level. DBM at individual level was defined as co-occurrence of stunting and overweight in the same child.Setting:Nationally representative surveys from ninety-three LMIC.Participants:A total of 825 633 children were studied.Results:DBM at national level was observed in five countries, whereas it would be expected to occur in eleven countries. Six countries did not present evidence of DBM at national level but did so in at least one wealth quintile. At individual level, thirty countries (32·3%) showed higher prevalence of DBM than would be expected, but most differences were small except for Syria, Azerbaijan, Albania and Egypt.Conclusions:The observed number of countries or socio-economic subgroups within countries with the DBM using recommended thresholds was below what would be expected by chance. However, individual-level analyses showed that one-third of countries presented higher prevalence of DBM than would be expected.