Heavy workloads have been central to the discussion of well-being. However, fluctuations in workloads have received sparse attention, even though transient and routine levels of workloads may have independent effects on well-being. The article assesses this, particularly focusing on the effort-recovery model of the workload/well-being relationship in which work-nonwork interference mediates this relationship. Using data from a weekly diary study of freelance or portfolio workers, multilevel analyses showed that both routine and transient levels of work to nonwork interference mediated the job demands-calmness relationship. Work to nonwork interference also negatively mediated the hours worked-calmness relationship, but the direct relationship between hours worked and calmness was a positive one so the mediating effect reduced this positive effect. In the case of enthusiasm, which was a second measure of well-being, there were no mediating effects but both routine and nonroutine levels of job demands and transient levels of hours worked were related to it. The study highlights the value of introducing the temporal dimension into the study of the work-nonwork interface, and the diversity of relationships that occur across different dimensions of workload and well-being.