Time-based pricing is used by two-sided platforms as a lever for regulating supply and demand during high-demand periods. In this paper, we study time-based pricing for peak demand periods in a static setting for an on-demand service platform where the peak period prices are announced in advance. We develop a Quadratic Programming (QP) model for computing the optimal price for high-demand slots considering the willingness-to-pay (WTP) of customers and expectation-to-be-paid (ETP) of professionals. We evaluate both homogeneous pricing (same price across high-demand slots) and heterogeneous pricing (price varying across high-demand slots) for maximising platform revenues. We analytically show that the optimal price always belongs to the set of WTP values of customers, irrespective of heterogeneous or homogeneous pricing across high-demand slots. We numerically show that the optimal solution obtained with enumeration is faster (up to 97%) than the QP for both homogeneous and heterogeneous cases and that social welfare for homogeneous pricing is higher compared to heterogeneous pricing for some instances. We also evaluate the impact of variation in demand-to-supply ratio, platform commission rate, and WTP, ETP distributions on platform revenue and social welfare and discuss key managerial insights.