Background: The Token Test (TT) is widely used to examine comprehension disorders in aphasic patients, but abilities other than language may affect a patient’s performance. This study aims to explore the correlation between the TT subtest performances and the performances in extra-linguistic cognitive areas in a cohort of patients from the Intensive Rehabilitation Post-Stroke (RIPS) study with a first, right hemisphere stroke and without aphasia, prospectively enrolled at admission to intensive inpatient post-acute rehabilitation. Methods: The patients were administered the TT (50-item version), the forward and backward digit span (DST), and the Montreal Cognitive Assessment (MoCA). Spearman’s partial correlations adjusted by age were used to evaluate the association between the number of errors in the TT and the other tests’ corrected scores. Results: Of the 37 patients enrolled in this study, 29.7% made 3–11 errors on the TT, 27.0% more than 11 errors, mostly in parts IV and V. The forward and backward digit span scores showed correlations with errors in part V of the TT (r = −0.408, p = 0.013; r = −0.307, p = 0.027). The errors in part IV of the TT presented a correlation with a forward digit span too (r = −0.394, p = 0.017). With respect to MoCA domains, executive functioning, and orientation were related to the TT part V errors (r = −0.468, p = 0.007; r = −0.499, p = 0.003). The orientation also correlated with the TT part III (r = −0.504, p = 0.002). Conclusion: Our findings show that the TT performances in patients with right hemisphere stroke and without aphasia are related to impairments in auditory–verbal span/auditory working memory mostly for TT scores on subpart V as measured by the DST and to executive function and orientation, as measured by the MoCA subtests.