This essay discusses methodological and ethical problems in the implementation and evaluation of home-based intervention programs for young children and their families. Part I notes the difficulties in (1) selecting an appropriate research design to evaluate a program (e.g., preventing sample bias), (2) specifying precisely the intervention method (e.g., insuring uniformity of treatment), (3) making decisions concerning which program effects to measure, (4) implementing the program in new settings (e.g., maintaining quality control) and (5) delivering the program in an ethically acceptable manner (e.g., avoiding subtle coercion in obtaining the consent of the target population). Part II briefly outlines recommended procedures to avoid the problems discussed in Part I. (BH)