Introduction and Aim: COVID-19 outbreak was declared as pandemic by WHO Director-General on 11th March 2020 in his opening remarks at the media briefing. The global population infected by Corona virus appeared to be responding at different levels in the first wave which warranted WHO to categorise the disease as mild, moderate, and severe. Haematological, biochemical, and radiological parameters played a crucial role in critically triaging and following-up disease progression. Amongst various laboratory parameters, this work aimed to identify the most specific marker in predicting disease severity. Materials and Methods: Blood samples from study population of 510 laboratory confirmed COVID-19 cases admitted in our hospital were selected. The patients who were classified as having mild, moderate, and severe disease were analysed for biochemical and haematological inflammatory markers. Results were analysed by ANOVA and post-hoc tests. ROC curves were derived to determine the cut-off values between severe and non-severe groups. Correlation between D-dimer and NLR was done by Pearson’s correlation. Results: Patients with co-morbidities were likely to develop severe complications which could lead to poor outcome. From ROC curves, we determine that NLR, with highest area under curve, is the best marker of disease severity. A significant positive correlation was found between D-dimer and NLR (p=0.000) across groups. Baseline cut-off values for D-dimer and NLR based to differentiate between severe and non-severe cases were 0.5 and 4.875 respectively. Conclusion: We conclude that baseline NLR is a simple and most useful tool that would assist clinicians in designing treatment strategies for a COVID-19 infected patient.