Mind wandering refers to engaging in task-unrelated thoughts and takes place almost 50% of the time (Killingsworth & Gilbert, 2010). Consequently, performance in the main task, like sensory information processing (Smallwood et al., 2008), is impaired. In the present study, we investigated whether these attentive lapses may also influence parameters of dual-task performance, e.g., between-task interference (backward-crosstalk) and multitasking efficiency when performing two visual-manual number categorization tasks. Dual-task trials were interspersed by random thought probes requiring participants to rate their focus of attention prior to the probe as on or off task using two different rating metrics in Experiment 1 (N = 39) and in Experiment 2 (N = 40). To examine whether probes of increased mind wandering affected the backward-crosstalk effect (BCE), we restricted the analysis to the last four trials before every thought probe. Results revealed that mind wandering grew larger across the experiment and affected overall performance, i.e., reaction times and accuracy, but did not influence the BCE in both experiments. In a second analysis, we investigated effects of mind wandering on the dual-task efficiency marker Time on Task (ToT, Reissland & Manzey, 2016). Indeed, ToT was longer for trials accompanied by higher mind wandering in both experiments. These results suggest that attentive lapses due to mind wandering appear to selectively reduce the overall efficiency of performing two tasks while leaving between-task interference unaffected.