To the Editor: We thank Drs. Lipman and Boyle for their interesting comments on our article. This work does indeed sound like exactly the kind of study we were trying to outline. The observations about difficulties recruiting people and drawing attention to the importance of the first contact will be very useful for others embarking on community studies. The admonition that an intervention may have unpredicted effects is also very well taken. We disagree about the final point, however. We argue instead that one of the purposes of research in community settings is to calibrate the intervention to the realistic possibility of its poststudy implementation. To use an exaggerated metaphor, if a poor community is having transportation problems and a researcher wishing to solve the problem brings in a fleet of limousines to demonstrate that the problem can be solved, this is a study hardly worth doing. Limousines are not likely ever to be available in this community. Documentation that if they exist, people will ride in them is not helpful.