BACKGROUND: Identification and treatment of routine tobacco use in medical practice is an effective intervention but is not used consistently. A study was conducted at HealthPartners, a large network-model health plan in Minnesota, to determine the effect of an outcomes recognition strategy that involved bonus funds and the rates at which network physicians document that tobacco users are identified and advised to quit. METHODS: Audits of 14,489 ambulatory patient records from 19-20 medical groups were conducted to determine the proportion of charts from each medical group that demonstrated identification of smoking status and counseling to encourage quitting at the most recent office visit in each year. RESULTS: Overall mean tobacco use identification increased from 49% +/- 7% (95% confidence interval [CI]) in 1996 to 73% +/- 7% in 1999 (p < .001), while advice to quit increased from 32% +/- 10% in 1996 to 53% +/- 10% CI in 1999 (p < .005). The number of medical groups with tobacco status identified at > 80% of visits and > 80% of tobacco users advised to quit increased from 0 in 1996 to 8 in 1999. DISCUSSION: Data feedback combined with a financial incentive appear to be an effective way for a health plan to improve physician compliance with the tobacco treatment guideline. Other health plans might consider similar reporting and incentive approaches to effectively engage medical group leadership and to improve the health of their members who use tobacco.