The nutritional status of under-five children often cited as a sensitive indicator of the household’s living standard and the economic condition; and also an important determinant of child survival. Although India has significantly achieved remarkable progress in reducing child malnutrition, progress toward reducing the number of malnourished children has been sluggish. Improvements in nutrition still represent a massive unfinished agenda. The objective of this study identified the place-specific spatial dependencies and heterogeneities in the associations between socioeconomic and demographic factors and nutritional status among under-five children in India using the fourth wave of National Family Health Survey (NFHS-4 2015–16) data for 640 districts. The dependent variables were stunting, wasting and underweight. Moran’s I and univariate LISA were used to confirm the spatial autocorrelation and clustering of nutritional status. Multivariate Ordinary least square (OLS), Geographically weighted regression (GWR), spatial (lag/error) models were employed to decrypt the determinants of under-five nutritional status. Overall, the prevalence and spatial clustering (Moran’s I statistics) of stunting, wasting and underweight were 38% (0.634), 21% (0.488) and 36% (0.721), respectively. GWR results disclosed that the relationships between the outcomes and its covariates were significantly place-specific and spatially clustering in terms of their respective magnitude, direction and strength. Regarding model performance and prediction accuracy, GWR better fit compared to traditional OLS models. The findings of the present study identified district-level nutritional conditions (hotspots) in India, where children are under severe risk of malnutrition can help health professionals, planners and policymakers in designing and implementing effective place-specific health policies to improve district-level under-five nutrition status in India.