This statewide evaluation of New York's adult drug courts finds modest reductions in re-arrest over a three year study period, with wide variation in impact across the 86 sites. In fact, while the most effective courts reduce re-arrest by up to 21 percent, the least effective courts increase re-arrest. Factors associated with more positive results include targeting a high-risk offender population; targeting a population with substantial legal exposure (e.g., felony as opposed to misdemeanor defendants); establishing a formal sanctions schedule; imposing sanctions for each noncompliant act; and including prosecutors and defense counsel as active members of the drug court process.