BACKGROUND: Dengue, a mosquito-borne viral infection, poses a rapidly growing burden, particularly in low- and middle-income countries. Without early identification of patients at risk of severe outcomes (dengue haemorrhagic fever, severe dengue, and plasma leakage- the latter typically occurring on days 5-7 of illness), untriaged admissions lead to hospital overcrowding and suboptimal care. METHODS: This nested case-control study compared early-stage plasma samples (within the first 96 hours of fever) from dengue patients with and without plasma leakage. Thirty-four potential biomarkers, selected through systematic review, were tested on a multiplex bead-based immunoassay platform. Subgroup analysis stratified patients by primary or secondary dengue infection. FINDINGS: A total of 228 patient samples (114 had plasma leakage) were tested. Elevated Vascular cell adhesion molecule-1 (OR:3.289, 95% CI: 1.090-9.926, p<
0.05), and Interleukin 33 receptor levels (OR: 2.677, 95% CI: 1.244-5.856, p<
0.05) were associated with an increased risk of plasma leakage while eotaxin-1 was associated with a decreased risk (OR: 0.166, 95% CI: 0.057-0.483, p<
0.05). When adjusted for prior dengue exposure, additional biomarkers (C-X-C motif chemokine 11, serum amyloid A) were also associated with plasma leakage. INTERPRETATION: Plasma leakage in dengue, being more objectively measurable than other severe outcomes, offers a reliable endpoint for biomarker studies. Identifying biomarkers that predict plasma leakage strengthens the evidence base in dengue research. These biomarkers could improve clinical assessment and patient care in dengue cases.