Coronavirus disease-19 (COVID-19) can, in severe cases, lead to cytokine-release syndrome owing to an excessive immune response. The release of different cytokines aggravates disease severity. IL-1β is a pro-inflammatory cytokine, while IL-10 is an anti-inflammatory cytokine, and both are involved in the human immune response to infection. This study aimed to determine whether serum levels of IL-1β and IL-10 and the ratio of the two over time in patients with COVID-19 could facilitate early identification of disease severity. An analytical, observational time-series design was employed. Fifty participants were enrolled between May and October 2020 and were divided into two groups-non-severe (n = 20), and severe (n = 30). IL-1β and IL-10 were analyzed using BD cytometric bead array sets. Association of the IL-1β:IL-10 ratio with COVID-19 severity was analyzed using a Mann-Whitney test and Fisher's exact test. Optimal cut-off values to predict disease severity were determined by Youden's index. In non-severe and severe groups, the median serum levels of IL-1β decreased on day 3 (1.72 ng/mL and 2.10 ng/mL, respectively), then increased on day 6 (2.05 ng/mL and 3.31 ng/mL, respectively). However, the median of IL-10 increased on day 3 (1.88 ng/mL and 2.30 ng/mL, respectively) and day 6 (2.02 ng/mL and 2.39 ng/mL, respectively). There was no significant association between the IL-1β:IL-10 ratio and COVID-19 severity at any time-point (p>0.05). The cutoff value of serum IL-10 between the two groups on days 0, 3, and 6 was 1.09 pg/mL (sensitivity: 66.6%; PPV: 71.4%), 2.11 pg/mL (sensitivity: 67.7%; PPV: 50.0%), and 2.08 pg/mL (sensitivity: 78.6%; PPV: 70.9%), respectively. The IL-1β:IL-10 ratio was not correlated to COVID-19 severity. However, owing to its high sensitivity, IL-10 may be a potential biomarker for disease severity in severe COVID-19.