This laboratory study tested a causal model of the effects of changes in perceptions of subjective workload and effort in relation to performance during stable and after shifts in task demand. Accordingly, this study addressed the need for within-person examinations of how individuals as a function of self-regulation respond to shifts in task demand. Participants were 198 university undergraduates who were trained to perform a computer game representing a complex decision-making environment. Subjective workload, subjective cognitive effort, and objective performance were concurrently measured at regular intervals (i.e., every 60 s) in five 10-minute trials, two of which involved a shift, either an increase or decrease, in task demand. Relationships between variables were examined using a longitudinal, multilevel approach suitable for disaggregating within-person (i.e., state) and between-person (i.e., trait) components. The proposed model reflecting inconsistent mediation was consistently supported when conditions involved stable task demands. Specifically, in trials involving stable demands, changes in subjective workload had positive indirect effects but stronger negative direct effects on performance. However, there was little support for inconsistent mediation in trials involving shifts in task demand. Rather, dynamic effects were observed as a function of the shift in task demands such that changes in subjective workload showed increasingly positive effects on performance after increases in task demand but increasingly negative effects after decreases in task demand. In general, this research demonstrated the need to account for indirect effects such as the volitional aspect of control (i.e., effort) and dynamic effects as a function of shifts in task demand when trying to understand the relationship between changes in subjective workload and performance.