The idea of aiding as cognitive support is to offer the user the knowledge he or she is missing. Recently, we developed a design method for aiding that is based on explicit requirements of the human problem solver. This proved to be able to supplement a lack of human knowledge in a statistical analysis task. In this article we extend the aiding concept to time-pressured tasks and we investigate whether aiding can supplement lack of knowledge and capacity under tasks with high mental loading, such as dealing with irregularities in process control. We developed a simulator of the workplace of a railway traffic controller with an aiding function for dealing with irregularities (e.g., a switch getting out of order). Application of the design method proved to be possible for this task. We then conducted an experiment to study effects of the aiding on task performance, mental effort, and learning under low and high task load conditions. Users of the simulator dealt better and faster with irregularities when the computer provided aiding. The higher the task load was, the larger this beneficial effect was. For theory about human-computer interaction, this research points to possible positive effects of aiding on performance and learning as a consequence of reducing cognitive demands.