Acute respiratory distress (ARD) is a global health concern due to its high rates of morbidity and mortality in children. This chapter aims at identifying early risk factors and generating a predictive model to predict baseline ARD in African children with severe malaria. This multi-centre retrospective analysis used secondary data of participants in ‘African Quinine-Artesunate Malaria Trial’ (AQUAMAT) that was conducted from 2005 to 2010, among children ( \({<}15\) years) who had been hospitalised for severe malaria. The prediction model of baseline ARD was developed using multivariable binary logistic regression model. A nomogram was constructed to visualise the predictive model. The Receiver Operating Characteristic (ROC) curve was plotted to evaluate the discriminative power of the predictive model. Classification tree analysis was done to classify patients at a higher or lower risk of developing baseline ARD. All variables were analysed at baseline at \(5\%\) significance level. Of the \(5,426\) children admitted with severe malaria, 867 (16.0%, CI \(15.0{-}7.0\%\) ) had baseline ARD. The multivariable binary logistic regression model revealed that the major predictors of baseline ARD were pneumonia (odds ratio (OR) 2.49, CI \(1.99{-}3.13\) , p-value \(<0.001\) ), severe acidosis (OR 2.49, CI 2.09–2.97, p-value \(<0.001\) ), if a patient is currently treated for chronic illness (OR 2.32, CI 1.05–5.14, p-value \(=0.038\) ), hyperparasitaemia (OR 1.96, CI 1.21–3.16, p-value \(= 0.006\) ), sepsis (OR 1.46, CI 1.18–1.82, p-value \(= 0.001\) ) and respiratory rate (per minute) (OR 1.03, CI 1.03–1.04, p-value \(<0.001\) ). On the other hand, convulsions \({>} 30\) minutes (OR 0.77, CI 0.63–0.93, p-value \(= 0.007\) ) and severe prostration (OR 0.69, CI 0.55–0.88, p-value \(=0.003\) ) were associated with lower odds of baseline ARD. The predictive model was valuable in predicting baseline ARD with overall correct classification of \(68.7\%\) and area under the ROC curve of 0.75 CI 0.73–0.77). Classification tree ranked pneumonia, severe acidosis, hyperparasitaemia, sepsis, respiratory rate as well as severe prostration as major conditions classifying a patient to be at high risk of developing baseline ARD. We developed a predictive model, nomogram and a classification tree incorporating demographic and clinical features to predict the ARD risk in children with severe malaria. The model showed a good predictive ability. These findings could help in early diagnosis, management and timely interventions provided to such patients.