Background Among patients with rheumatoid arthritis (RA), smoking increases risk of severe RA and pulmonary and cardiovascular disease. Despite this, little is known about smoking cessation counseling by rheumatologists. Objectives We examined predictors of tobacco counseling in RA patients who smoke including the effect of perceived RA control. We hypothesized that patients with controlled RA would receive more counseling according to the competing demands model, which explains that preventive care gaps occur as a result of competing provider, patient, and clinic factors. Methods This secondary data analysis involved RA patients with an additional cardiovascular disease risk factor identified in an academic medical center 2004–2011. Trained abstractors assessed documented smoking counseling and rheumatologists’ impression of RA control in clinic notes. We used multivariable logistic regression to predict having received smoking cessation counseling, including sociodemographics and comorbidity in models. Results We abstracted 3396 RA visits, including 360 visits (10%) with active smokers. Perceived controlled RA was present in 31% of visits involving smokers (39% in nonsmokers). Beyond nurse documentation, providers documented smoking status in 39% of visit notes with smokers and smoking cessation counseling in 10%. Visits with controlled versus active RA were less likely to include counseling (odds ratio, 0.3; confidence interval, 0.1–0.97). Counseling was more likely in visits with prevalent cardiovascular, pulmonary, and psychiatric disease, but decreased with obesity. Conclusions Smoking cessation counseling was documented in 10% of visits and was less likely when RA was controlled. Given smoking’s impact on RA and long-term outcomes, systematic cessation counseling efforts are needed.