Offshore supply vessel (OSV) collisions have been identified as the most frequent type of collision accidents in the offshore oil/gas or wind turbine industries. Quantitative risk assessment (QRA) is an efficient method for evaluating the collision risk to an offshore installation. In-depth information on collision load parameters, such as incoming ship velocity and impact location, is considered prerequisite for determining the consequences of collisions with accuracy. Thus, the aim of this study is to provide a new probabilistic method for determining collision design loads. Each input parameter is treated by a probability density function, and a set of 50 prospective collision scenarios is generated using Latin hyper cube sampling (LHS) technique. Numerical computations of ship motions are performed to obtain collision load parameters. The probabilistic characteristics of the parameters using a goodness of fit test and an interval study are carried out, and best-fit probability density functions and the exceedance curves are established. A case study of the proposed method is demonstrated using a hypothetical OSV and an offshore jacket structure located in a hypothetical oceanic region. The details of the computations are documented, and the findings of the study are discussed.