While in service, marine risers are subjected to various types of loads, such as axial tension and bending moments arising from waves, winds, and currents in association with the motions of the offshore platform. They are also subjected to internal and external pressure loads caused by internal flows and external water pressure. The characteristics of the loads on a marine riser are essentially probabilistic in nature, as they involve several uncertainties associated with random variables. The aim of this study is to quantify the probabilistic distribution of loads on a marine riser to aid determination of the nominal values of design loads. Two methods are investigated. The first is to select a set of credible scenarios in association with site-specific metocean data on an offshore platform, and then perform dynamic riser analysis to describe the probabilistic distribution of the loads. The second is to calculate a metamodel to predict the loads as a function of multiple input variables, a method that can also characterize the probabilistic load distribution by running a Monte Carlo simulation. Both approaches are compared via a numerical example of a marine drilling riser in ultra-deep water; the results show that metamodel-based method is the most appropriate to describe neatly the loads at low probability of exceedance. The characteristics of the loads on a marine riser are observed to be highly random and significantly affected by environmental and functional conditions. Hence, the design loads must be determined by considering the marginal probabilistic density function of all such parameters.