Objective: Little is known about the relationship between financial incentives and provider behavior. We examined restoration rates to determine the impact of an incentive system implemented at a large, multi-clinic dental group in 2009 targeting remineralization in low risk patients. Method: All newly diagnosed early stage caries (F80, F81, and F82 codes) and co-occurring findings for 35 providers from the years 2008 and 2009 were identified. Findings were classified into three treatment categories based upon treatment codes occurring within six months of the initial finding date: Fluoride, Remineralization, and Restoration. There were a large number of findings without identifiable treatments, but rates of unidentified treatment did not differ by study year (p=.35) and were dropped. Generalized linear models adjusting for clinic and patient factors compared restoration rates and provider-specific change from 2008 to 2009 coinciding with implementation of the incentive system. Result: 14,367 and 16,752 early stage caries were identified in 2008 and 2009, respectively. Overall, dental group use of restorations dropped 15.8% (74.3%-2008, 62.6%-2009; p<.01); however, there was considerable variation across providers. 29 providers reduced their use of restorations with the largest reduction being 56.7% (86.1%-2008, 35.9%-2009; p<.01). The remaining six providers increased their use of restorations with the largest increase being 4.5% (63.8%-2008, 67.7%-2009; p=.11). Across all 35 providers, 2008 restorations rates were negatively correlated with the change in restoration rate (r=-.19) indicating providers with the highest baseline use of restorations tended to have the largest change in their use of restorations. Conclusion: Incentive programs promoting use of non-invasive treatment of early caries appear effective at changing care in a desired direction; however, design of these programs must account for provider-specific baseline patterns of care and potential for individual improvement.