Recent research suggests that marital quality predicts the survival of patients with heart failure (HF), and it is hypothesized that a communal orientation to coping marked by first-person plural pronoun use (we talk) may be a factor in this. During a home interview, 57 HF patients (46 men and 16 women) and their spouses discussed how they coped with the patients’ health problems. Analysis of pronoun counts from both partners revealed that we talk by the spouse, but not the patient, independently predicted positive change in the patient’s HF symptoms and general health over the next 6 months and did so better than direct self-report measures of marital quality and the communal coping construct. We talk by the patient and spouse did not correlate, however, and gender had no apparent moderating effects on how pronoun use predicted health change. The results highlight the utility of automatic text analysis in couple-interaction research and provide further evidence that looking beyond the patient can improve prediction of health outcomes.