Human body size, shape, stature, joint range of motion, joint strength, and other factors vary from one person to another. Even for a single person, anthropometric data, such as weights and joint strengths, change with time. Due to this variability, different people adapt different postures to perform the same reach task within a vehicle. Even for the same person and reach task, postures will vary with time. Therefore, it is important to consider the reliability of achieving a reach task within a vehicle to create a better design for vehicle controls, enhance driver safety, and increase the level of accommodation for all types of drivers. In this study, we will present a reliability/probability approach to gain insights into driver reach tasks with uncertainty. Sensitivity levels are found to determine the importance of each joint to the reach tasks. A digital human upper body model with 21 degrees of freedom (DOFs) is introduced to demonstrate the probability approach. Seven different reach tasks for points to represent the seatbelt, visor, rearview mirror, gearshift, parking brake, radio tuner, and glove box are used to compare the probabilities of achieving these tasks and the sensitivities due to joint angle uncertainty. The results show that the importance of each joint is dependent on the reach task and that joint angle uncertainty affects reach probabilities for all tasks. For each reach task, the probabilities of failure and sensitivities of the joints are different. This indicates that it is necessary to incorporate these uncertainties into posture prediction. These results will help designers to design vehicle layouts to ensure that all drivers can reach buttons and controls within a vehicle.