Little research has been conducted on the effectiveness of training strategies for child care providers. The current study used an experimental intent-to-treat design to measure the impact of an established intervention, the on-site caregiver training component of the Program for Infant/Toddler Care (PITC), on child development and child care program quality. The PITC was developed by WestEd in 1985, in partnership with the California Department of Education. Over the next 25 years, more than 1,500 early childhood trainers across 30 states became PITC-certified trainers. More than 1,000 Early Head Start trainers have also been trained by the PITC. This study is the first rigorous effectiveness trial of the on-site caregiver training component of PITC. It was implemented over 2007-2010 in six Southern California counties and four Arizona counties. The study sample of 251 child care programs included 92 child care centers and 159 licensed family child care homes, and the sample of 936 children included an average of eight children per center and between one and two children per family child care home. The primary questions focus on child outcomes: (1) What is the impact of the PITC on a composite measure of children's cognitive and language skills, at least 6 months after its full delivery to the children's child care programs (within an average of 23 months after random assignment)?; and (2) What is the impact of the PITC on a composite measure of children's social and behavioral skills, at least 6 months after its full delivery to the children's child care programs (within an average of 23 months after random assignment)? The secondary questions focus on child care quality: (1) What is the impact of the PITC on global child care quality at least 4 months after the PITC ends (within an average of 21 months after random assignment)?; and (2) What is the impact of the PITC on a composite measure of the quality of child care programs' staff-child interactions at least 4 months after the PITC ends (within an average of 21 months after random assignment)? The primary findings are: (1) The PITC did not have a statistically significant effect on a composite measure of children's cognitive/language scores, measured approximately 6 months (on average) after it ended; and (2) The PITC did not have a statistically significant effect on children's composite behavior scores, measured at 6 months after it ended. Sensitivity analyses, conducted with two alternative approaches to missing data treatment, had results consistent with these findings. Secondary research questions addressed the effects of the PITC on child care program quality at, on average, four months after the intervention ended. These estimates also found no significant effects. Findings of this analysis are: (1) The PITC did not have a statistically significant effect on global program quality, as measured by trained observers administering the ITR and the FCCERS-R; (2) The PITC did not have a statistically significant effect on staff-child interactions, a composite measure incorporating interactions items from the environment rating scales and from the PITC-PARS. Results of sensitivity analyses were consistent these findings. Analysis of implementation found that, in many child care programs, the intervention was not fully implemented or was not implemented with full participation: Of the 124 child care programs assigned to the treatment group, 11 decided not to participate before receiving any training, and 6 dropped midcourse. In only 59.4 percent of participating family child care homes did at least one caregiver receive the benchmark 56 hours of training, and in 41.9 percent of child care centers, four or more caregivers (the minimum number of participants, plus the director, required for PITC delivery) received at least 56 hours of training. Of children in the treatment sample, 17 percent received no exposure to the PITC, either because they left their original child care programs before start-up or because their programs were among the 11 that declined the intervention after random assignment. The PITC incorporates a number of the features that preliminary research and expert opinion in the field suggest are most likely to have a positive effect: focus on relationships, on-site consultation, opportunities for assessment and feedback, and application to practice. However, this study finds no positive main effects and also underscores the difficulties of sustaining participation in an intensive, long-term intervention in a large number of community child care settings across geographically dispersed locations. More research on the PITC and other training interventions is needed for fuller examination of both implementation and impacts. Increased understanding of the "transfer" between training strategies, program quality, and child development would inform improved child care training design implementation. Appended are: (1) Sample Power Estimates; (2) Child Care Provider Screening Interview (California); (3) Method of Random Assignment; (4) Random Assignment Cohorts and Strata; (5) Follow-Up Data Collection Intervals, by Experimental Condition; (6) Details of Study Measures; (7) Zero-Order Correlations among Variables Used in the Impact Analysis; (8) Caregiver-Child Quality Interaction Composite: Factor Analysis; (9) Training and Reliability-Checking of Field Staff; (10) Detailed Response Rates and Reasons for Nonresponse, by Experimental Condition; (11) Teacher Sample Turnover and Response Rates; (12) Additional Sample Equivalence Tables; (13) Child Care Licensing Regulations in the Study States; (14) Program for Infant/Toddler Care Professional Growth Incentives for Child Care Center Directors, Child Care Staff, and Family Child Care Providers (California); (15) Sensitivity of Impact Estimates to Alternative Model Specifications; and (16) Additional Subgroup Analyses of Child Mobility. (Contains 9 figures, 78 tables and 14 footnotes.)