This paper examines how rainfall fluctuations in the first year of life affect children's anthropometric status in Papua New Guinea (PNG), an understudied context despite having one of the world's highest stunting rates. It draws on a single cross-sectional survey, the 2023 PNG Rural Household Survey (IFPRI), covering 2,699 households in 270 communities across 14 of PNG's 22 provincial-level divisions. Anthropometric measurements were taken for every child under five (1,334 children in 1,008 households) and converted into WHO height-for-age (HAZ), weight-for-age (WAZ) and weight-for-height (WHZ) z-scores, plus binary stunted, wasted and underweight indicators. Rainfall comes from Multi-Source Weather (MSWX), a bias-corrected global gridded dataset at 10 km resolution. For each community and calendar month, the authors estimate a rainfall trend over the thirty years preceding the survey (May 1993–April 2023) and express observed rainfall as a standardised deviation from that trend; each child's rainfall shock is the average of these deviations over the twelve months following birth. Anthropometric outcomes are then regressed on this shock and on its interaction with a male dummy, controlling for child age and for community, birth-year and month-of-birth fixed effects, with standard errors clustered at community level. 

Rainfall shocks have no significant effect on the binary indicators: shortfalls do not push more children over the stunting, wasting or underweight thresholds. They do, however, slow growth as measured by the continuous z-scores — and only for girls, with no effect for boys. A one-standard-deviation rainfall shortfall lowers girls' z-scores by around 0.15, roughly a tenth of the average height-for-age deficit in the sample. They read this gender gap as consistent with boys being favoured in the intra-household allocation of resources, particularly food, when adverse rainfall tightens household budgets. The mechanism is inferred, not observed, as the data do not allow it to be tested.