High urbanization rates in Latin America are accompanied by an increase in women’s participation in the labor force and the number of households headed by single mothers. Reliable and affordable childcare alternatives are thus becoming increasingly important in urban areas. The Hogares Comunitarios Program (HCP), established in Guatemala City in 1991, was a direct response to the increasing need of poor urban dwellers for substitute childcare. This government-sponsored pilot program was designed as a strategy to alleviate poverty by providing working parents with low-cost, quality childcare within their community. This paper presents preliminary findings from an evaluation of the HCP carried out in 1998 in urban slums of Guatemala City. The evaluation included both an operations (or process) evaluation and an impact evaluation. Key findings of the operations evaluation are summarized, and preliminary findings of the impact evaluation on children’s dietary intakes are presented. Aspects related to the targeting, coverage, and cost of the program are also discussed, and the patterns of childcare use by nonbeneficiary households and their costs are described. PROGRAM BACKGROUND In the HCP model, a group of parents elects a neighborhood woman to act as a “caretaker” mother. This mother then receives and cares for up to 10 children in her home, 12 hours a day, five days a week. During their stay in the caretaker’s home (herewith referred to as the hogar), the children receive care and affection, hygiene, early child stimulation, and food. The program provides initial training for the caretaker mothers and furniture, cooking equipment, and supplies for 10 children. On a monthly basis the program gives approximately $0.60 per child per day to the caretaker for food, fuel, and educational material. The program also gives the caretaker an “incentive” of $3 per child attended per month, which is complemented by a $5 per child contribution from the parents. OPERATIONAL EVALUATION The operational evaluation had three objectives: (1) to review and evaluate the effectiveness of implementation; (2) to assess the quality of the services provided by the caretakers; and (3) to evaluate the level of satisfaction and the attitudes of the program’s main implementers (caretakers and their supervisors) and main users (the beneficiary parents). Both qualitative and quantitative methods were used, including semistructured interviews with caretaker mothers, eight-hour observations in the hogares, and focus groups with caretakers, beneficiary parents, and field supervisors. The evaluation was carried out in 206 hogares operating in three zones of Guatemala City. Findings show that the program is generally well-designed and is operating effectively. Delays in cash transfers for food, the insufficient amount of the transfer, and the lack of participation of parents were the key operational constraints identified. The quality of services provided by the caretakers with respect to hygiene, safety, and their interaction with children was good overall but varied significantly between hogares. Caretakers consistently failed to allocate the required amount of time to educational activities, largely because of time constraints, but also because they did not feel adequately trained, motivated, and remunerated. Caretaker mothers were generally grateful to the program for the opportunity to work while taking care of their children (or grandchildren) at the same time. Beneficiary parents were extremely positive about the program. They were appreciative of the caretakers and of the program for the assistance received, and indicated that the program was affordable. They suggested the addition of Saturday care and an increased emphasis on preventive and curative health care. Most of the recommendations to correct some of the constraints identified by this evaluation were accepted by the new administration that took over the program in 2000. Concrete actions to address and adopt the recommendations were included in the new four-year plan. These included increasing the amount of the cash transfers, strengthening preventive and curative health services, hiring educators to ease the time constraints on caretaker mothers, and strengthening human resource through additional training. IMPACT EVALUATION The impact evaluation was carried out in one zone of Guatemala City and included (1) a case-control design of approximately 250 beneficiary children matched with control children of the same age and neighborhood, and whose mothers also worked outside the home, and (2) a random sample of approximately 1,400 households with children 0–7 years of age. The main objective was to assess the impact of the program on children’s dietary intakes, maternal wages and employment conditions, household expenditure patterns, and older siblings’ school attendance. Only preliminary findings of the impact of the program on children’s dietary intakes are presented here. The purpose of the random sample was to examine aspects of targeting, coverage, patterns of use of other types of childcare and their cost, and to address the issue of whether the program affected women’s labor force participation. The program appears to be reaching its targeted population, i.e., families of working parents with poor resources and particularly families where mothers are the main income generator. Beneficiary mothers are more likely to have a salaried (and possibly more stable) employment than mothers who use other childcare arrangements, which results in higher wages and a larger number of employment benefits. Among nonbeneficiary families, the most commonly used childcare arrangements involved household members or extended family members. Even compared to these informal alternatives, the HCP was one of the lowest cost alternatives, ranking second after resident household members. Nonresident relatives were more costly than the HCP, as were neighbors, other private arrangements, and formal childcare. The low coverage of the program (only 3 percent of working mothers in the random sample used the program) seems to result from lack of supply rather than low demand. The program is having a significant and positive impact on children’s nutrient intake and dietary diversity: children participating in the program consume, on average, 20 percent more energy, proteins, and iron, and 50 percent more vitamin A than do control children. Moreover, a greater proportion of the key micronutrients (iron and vitamin A) consumed by beneficiary children is from animal products, and thus are more bioavailable (more easily absorbed and used by the body). Because the home diet of beneficiary children was also slightly more nutritious compared to control children, the net nutritional impact of the program is positive and significant. CONCLUSIONS The government-sponsored HCP in Guatemala provides affordable and good quality childcare for extended hours, thereby providing needed support to vulnerable urban households, namely single mothers. Expansion and continued strengthening of this type of program could significantly contribute to reducing urban poverty, food insecurity, and childhood malnutrition.