Mental workload is the amount of mental effort necessary to perform a task or set of tasks and is an important factor for predicting performance. The research described here included four experiments each of which investigated a different task type with an independent variable of the number of tasks performed at once. The four experiments acted as follow-up experiments to those described in Cassenti, Kelley, and Carlson (2010) with changes in the time available to respond (TAR) to each task. The findings suggest that a modeling system should predict fewer errors with more TAR, less effortful tasks, and fewer tasks. Together, these experiments illuminate a way forward to constructing performance-prediction algorithms based on number of tasks, task types, and TAR. We propose that based on these results and others a plug-in could be added to a human performance modeling system such as the IMproved Performance Research INtegration Tool (IMPRINT) to estimate performance given the number, type, and TAR of tasks. An example of how algorithms that populate this plug-in may be implemented is illustrated and equations for a host of task types are provided.