We find that the entitlement to child care subsidies and policies associated with the Starting RIght initiative in Rhode Island significantly increased the likelihood that current and former welfare recipients would use child care subsidies and significantly increased the availability of formal child care. In addition, these policy changes increased work among cash assistance and non-cash-assistance recipients and encouraged cash recipients to leave welfare for work. The most powerful impact of the changes in child care policies was on families that left welfare (former cash recipients) and worked at least 20 hours per week. These policy changes had less effect on families on cash assistance participating in some activity other than work. We were not able to assess the impact of RI policy changes on families who were never on cash assistance. However, the large increase in the number of such families receiving child care subsidies suggests that the impact may have been substantial. To discern the impacts of Starting RIght, we constructed a longitudinal database for the period May 1996 (a year before RI's welfare reform) to April 2000. This database encompassed information from many sources, including DHS administrative data for families enrolled in the child care subsidy program and those currently and formerly enrolled in the cash-assistance program, information on employment records from the administrative files of the Unemployment Insurance (UI) program, and, for all townships in RI, information on the availability of formal child care and education and on the labor market. Using economic modeling and the multinomial logit estimation technique, we find that Rhode Island's changes in child care policies significantly increased the likelihood that working current and former cash-assistance recipients living in one of the core cities would use child care subsidies. Policy changes had both a direct effect on the take-up rate for child care subsidies and an indirect effect through their impact on the availability of care. Considering both direct and indirect effects, we estimate that the child care policy changes associated with welfare reform and with Starting RIght increased by between 12% and 23% the probability that a working former recipient family would use child care subsidies and increased by between 11% and 13% the probability that a working recipient family would use child care subsidies. Our best estimate is that between 65% and 76% of eligible former recipients and between 67% and 69% of working recipients living in one of the core cities were using child care subsidies by the second quarter of 2000. By increasing reimbursement rates and by subsidizing health insurance for family child care providers and for center employees serving subsidized children, Rhode Island was able to substantially increase the availability of formal child care, particularly in the core cities where most low-income families live. The availability of preschool care at centers expanded throughout the state, increasing particularly rapidly in Central Falls and Pawtucket. The availability of school-age care expanded rapidly throughout the state, with the highest rate of expansion occurring in the core cities after Starting Right. The availability of family child care expanded moderately in 1997 and in 1998 after subsidized health insurance coverage was made available to family care providers serving subsidized children, and it expanded at a much faster pace in 1999 and in 2000 after Starting RIght. We found substantial evidence that Rhode Island's child care policy changes served as an important support that allowed low-income families to find employment and to leave the cash-assistance program. (author abstract)