Objectives: Evidence supports the view that reductions in cognitive hyperarousal contribute substantially to improved sleep outcomes following cognitive and behavioral interventions for insomnia disorder. Assuming an inverted-u relationship between arousal and performance, a theoretical possibility, supported by limited empirical data, is that the same mediating processes could negatively impact aspects of psychomotor performance, reducing speed on tests of reaction time. Participants: Sedentary participants (mean age = 59.8; SD = 9.46) meeting research diagnostic criteria for insomnia were randomized to either an exercise intervention of ≥150 min of moderate-intensity activity per week (n = 20), or a wait-list control group (n = 21). Of these, n = 17 intervention and n = 18 control participants completed 6-month follow-up assessments. Methods: Digit span, and simple and complex vigilance task performance was assessed using a computerized protocol at baseline and 6-month follow-up. Dependent variables included digit span, simple reaction time (SRT), complex reaction time (CRT), false positive responses, number of lapses, and SRT/CRT ratio (indicative of the magnitude of difference between simple and complex RT performance). The primary clinical outcome was Insomnia Severity Index (ISI) score. Results: In comparisons of baseline to follow-up change, ISI scores showed clinically significant improvement in the intervention group at 6-month follow-up (F (8,26) = 5.16; P = 0.03). Baseline vigilance performance was equivalent in both groups. At 6-month follow-up, however, the intervention group showed significantly slower simple reaction time F(4,30) = 10.25, p < 0.01, and a significantly decreased SRT/CRT ratio (F(4,30) = 13.22, p < 0.01). Conclusions: Among people meeting diagnostic criteria for insomnia, beneficial sleep outcomes following successful behavioral interventions may, under some circumstances, come at the cost of slower psychomotor performance.