The impact of cerebral malaria on the transcriptional profiles of cerebral tissue is difficult to study using non-invasive approaches. We isolated plasma extracellular vesicles (EVs) from patients with cerebral malaria and community controls and sequenced their RNA content. Deconvolution of the tissue origins of the EV-RNA revealed that EVs from cerebral malaria patients are predominantly enriched in transcripts of brain origin. Next, we used manifold learning on the EV-RNAseq data to determine pseudotime against the community control samples as the baseline reference. We found that neuronal transcripts in plasma EVs decreased as pseudotime progressed, while transcripts of glial, endothelial, and immune cell origins increased over pseudotime. Pseudotime was associated with clinicopathological parameters of disease severity, including retinopathy, metabolic acidosis, respiratory rate, anaemia, malnutrition, depth of unconsciousness and death. Plasma EVs further provided evidence of platelet activation, TNF signalling, neurotrophin signalling, long-term potentiation and glutamatergic signalling during late disease stages of cerebral malaria. The transcriptional responses of cerebral tissue in cerebral malaria can be studied non-invasively using EVs circulating in peripheral blood.