BACKGROUND: International guidelines recommend that primary care clinicians recognize obesity and offer treatment opportunistically, but there is little evidence on how clinicians can discuss weight and offer treatment in ways that are well received and effective. OBJECTIVE: To examine relationships between language used in the clinical visit and patient weight loss. DESIGN: Mixed-methods cohort study. SETTING: 38 primary care clinics in England participating in the Brief Intervention for Weight Loss trial. PARTICIPANTS: 246 patients with obesity seen by 87 general practitioners randomly sampled from the intervention group of the randomized clinical trial. MEASUREMENTS: Conversation analysis of recorded discussions between 246 patients with obesity and 87 clinicians regarding referral to a 12-week behavioral weight management program offered as part of the randomized clinical trial. Clinicians' interactional approaches were identified and their association with patient weight loss at 12 months (primary outcome) was examined. Secondary outcomes included patients' agreement to attend weight management, attendance, loss of 5% body weight, actions taken to lose weight, and postvisit satisfaction. RESULTS: Three interactional approaches were identified on the basis of clinicians' linguistic and paralinguistic practices: creating a sense of referrals as "good news" related to the opportunity of the referral (n = 62); "bad news," focusing on the harms of obesity (n = 82); or neutral (n = 102). Outcome data were missing from 57 participants, so weighted analyses were done to adjust for missingness. Relative to neutral news, good news was associated with increased agreement to attend the program (adjusted risk difference, 0.25 [95% CI, 0.15 to 0.35]), increased attendance (adjusted risk difference, 0.45 [CI, 0.34 to 0.56]), and weight change (adjusted difference, -3.60 [CI, -6.58 to -0.62]). There was no evidence of differences in mean weight change comparing bad and neutral news, and no evidence of differences in patient satisfaction across all 3 approaches. LIMITATIONS: Data were audio only, so body language and nonverbal cues could not be assessed. There is potential for selection bias and residual confounding. CONCLUSION: When raising the topic of excess weight in clinical visits, presenting weight loss treatment as a positive opportunity is associated with greater uptake of treatment and greater weight loss. PRIMARY FUNDING SOURCE: National Institute for Health and Care Research School for Primary Care Research and the Foundation for the Sociology of Health and Illness.