Since 2004, regional autonomy in Indonesia has not contributed to corruption control in the expected way. In contrast, the shifting of power from central to local government has created decentralized corruption practices at the local level, with local legislators, executives, law enforcement authorities and new oligarchs involved in various corruption events at one time or another in most regions. My research consists of a comparative analysis of various anti-corruption responses of three Indonesian city governments (Jakarta, Surabaya, and Medan) and completes previous findings on the effectiveness of anti-corruption programs. Some factors (transparency, accountability, leadership, law enforcement, and bureaucratic reform) have been chosen in this study to diagnose their role in both success and failure of anti-corruption programs in these cities. The relationship between factors is then utilized to create the city's strategic anti-corruption programs. It appears that leadership is the most influential element in the city image and is the dominant factor in changing a city's performance, either in a positive or a negative manner. A leader who can serve as a role model, has a vision, and implements his commitment to anti-corruption programs can transform a city more effectively and affect other contributing factors, as it happened in North Jakarta and Surabaya. By contrast, with a weak leader a city will drift to the bottom line of corruption and its performance will be greatly affected as it was found in Medan City. Although this study finds that leadership is the dominant influence in either encouraging or discouraging the transformation of a city, it is not the only factor that determines a city's performance in the fight against corruption. Other balancing factors such as intervention by the KPK and public engagement are also necessary to achieve success in an anti-corruption program. Although intervention by the KPK just appeared in a small area of Medan City, it shows a promising anti-corruption approach in a city with poor leadership. Moreover, this study also reveals that public engagement has impacted positive public perception significantly as shown in the case of North Jakarta City; conversely, lack of public engagement has been a factor in Medan City's failed anti-corruption program. Based on the dominant factors that have been identified, this research develops an appropriate strategy using a policy evaluation approach to create a practical guide regardless of whether there is good or bad leadership in a given city. This method will halt the city's corruption cycle which so far has mostly depended on a leader's performance and introduce a new approach that interconnects leadership with other factors. By developing evaluation effectiveness criteria, determining strategies, planning, and reducing calculation errors in achieving a city's anti-corruption goals, this approach will bridge theory and practice to maximize anti-corruption efforts in a city. This research is expected to be a useful reference for stakeholders, especially the government, as a blueprint to prevent corruption by considering several important steps and guidance in the effort to eradicate corruption in cities, especially in Indonesia.