This paper addresses the question of early discrepancies in illness cognitions and mood between 72 stroke patients and their informal careers, and examines whether discrepancies are predictive longitudinally of the patient or career mood. 72 couples who acted as controls in a randomized controlled trial of a stroke workbook-based intervention were assessed approximately 3 weeks post-discharge (T1), 7 weeks later (T2), and finally 5 months thereafter (approx 190 days post-discharge, T3). Participants were assessed in terms of mood (HADS); perceived control over the recovery (RLOC), recovery confidence, and perceived patient disability. Discrepant responses were calculated by subtracting career scores from patient scores. Descriptive results show that at the stage of T1, careers identify patients to be more disabling than they see themselves to be, careers are supplementary concerned and not as much of disheartened than patients, and patients have better improvement assurance. Supposed organize did not change between groups. Differences within married couples in emotions following a stroke, or in recovery beliefs, have differential effects. It is proposed that dyadic or individual coping, and premorbid marital interaction may mediate potential negative effects of discrepant responses to acute stroke.