Availability of geo-referenced data has increased applications of spatially explicit models to understand important health problems in developing countries. This study aims to investigate joint and disease-specific spatial clusters of fever and diarrhoea at a highly disaggregate level, while simultaneously estimating the influence of other covariates. Using the 2000 MalawiDHS, a logistic model was fitted with spatial random effects partitioned into shared and specific effects. Results indicated that the shared area-specific effects were persistently high in the central and southern regions. Fever-specific effects were high along the lakeshore areas of the country, while diarrhoea-specific effects were excessive in the central region and south-eastern zones of the country. The prevalence of fever and diarrhoea was also associated with individual, familial and community risk factors. Our findings present an opportunity for an integrated disease control approach for reducing childhood morbidity and mortality.