Background/Aim: Childhood anaemia continues to persist as a prominent nutritional disease and a public health challenge in India despite several initiatives by the Government of India. This study aimed to identify predictors and regional disparities for targeted interventions. Methods: This study utilised data from a nationally representative cross-sectional survey from the fifth round of the National Family Health Survey (NFHS-5), encompassing 177,695 children aged 6-59 months across 707 districts and 36 states and union territories of India. It employed multivariate logistic regression and spatial analysis at district levels to examine socio-demographic predictors and spatial patterns of childhood anaemia in the country. Result: Multivariate logistic results revealed, women aged 15-19 were 2.43 times more likely to have an anaemic child compared to those aged 35-49 and uneducated mothers had a 29 % higher likelihood of having an anaemic child. There was positive spatial autocorrelation (Moran's I value = 0.579) at the district level in India, with 108 identified hotspots in regions including Jammu and Kashmir, Ladakh, Gujarat, Maharashtra, Telangana, Uttar Pradesh, Rajasthan, Jharkhand, Chhattisgarh and Bihar. The spatial error model (SEM) indicated that mother's anaemia (0.53) and maternal education (0.23) were key predictors of child anaemia in India. Conclusion: The study findings provide valuable understanding regarding the socio-demographic predictors associated with childhood anaemia such as adolescent motherhood, low education, lack of media exposure, higher birth order and rural residence. Also, the spatial study provides the spatial heterogeneity of childhood anaemia at the district level and advocates more attention toward hotspot regions in the country.