Summarizes final evaluation (XD-ABJ-177-A) of a project (11/91-9/93), implemented by CARE in northwestern Guatemala, to construct water supply structures and latrines and provide health education and training in water system maintenance. The evaluation focused on (1) three activities which CARE added as pilot efforts (a behavior-based monitoring system, partial cost-recovery, and microwatershed management); and (2) health education in 12 communities where water/sanitation systems had already been installed. Findings were as follows. (1) In the behavior-based monitoring system, community volunteers conduct monthly home visits to track key health and sanitation indicators with the help of a pictorial checklist. The program has been well-received and has enabled volunteers to focus attention on hygienic practices. Recommendations for improving the program include making the home visits bi-monthly, reducing the number of indicators from 22 to 10, and engaging a professional illustrator to improve the clarity of the pictures. (2) The cost-recovery system was also well-received by participating communities; the target of 30% recovery of construction costs was found to be reasonable, and CARE is cautioned against making exceptions either for communities or households as this could cause conflict. If exceptions are to be made for poorer families or individuals, that decision should be left to the community. CARE should also try to collect as much money as possible prior to completion of the system, when incentive to pay is probably at its highest. (3) Microwatershed management activities suffered from both a late start and a lack of specific objectives, although they have helped to raise awareness of the consequences of deforestation; so far, however, the main activity has been the establishment of tree nurseries of fast-growing local species; it is recommended that the emphasis be shifted from agroforestry to educating communities about the link between their water supply and the health of the watershed. (4) Follow-up health education was also late in starting; implementation occurred only during the last 9 months of the project. In addressing only health education, the component ignored other important issues such as systems operations and the role of water committees. CARE should consider conducting extended follow-up with a small staff, eventually transferring responsibility to a local organization. All project elements, not just the educational ones, should be addressed, and educational emphasis should be placed on practices which prove most difficult to change. Lessons learned included the following. (1) Health/hygiene education is most effective when community participation is at its height, i.e., when water/sanitation systems are being constructed. (2) The follow-up activities improved hygiene practices, although in this case the improvement was small due to high baseline indicators. A good follow-up period increases the likelihood of sustainability, both of water systems and hygiene practices. (3) Specific, measurable objectives should be established for the environmental education activities. (4) The real value of a monitoring system is to provide support and feedback to community volunteers, as well as periodically to evaluate results. (5) Cost-recovery is acceptable to rural communities if it is within their financial capability; increasing communities' sense of ownership of water systems can improve system maintenance and operation; however, if poorer communities refuse to pay it may be necessary to consider making exceptions.