This paper examines the major problems which confront researchers who wish to study sensitive topics in family relations. The major obstacles typically encountered are (1) locating subjects; (2) getting cooperation; and (3) obtaining valid and reliable data. A second section of the paper discusses methods of overcoming these obstacles. One subsection lists and discusses sampling techniques which have been proposed and/or implemented to study such "taboo" topics as sexual behavior, homosexuality, family violence and child abuse. The next subsection discusses methods of data collection. Standard methods such as observations and interviews tend to require considerable work if researchers are to get the cooperation of the subjects. Techniques such as the funneling technique, the random response technique, projective techniques, and experimental designs are reviewed in this section. The last subsection examines problems of validity and reliability. The third section examines additional problems in sensitive issue research such as contingencies posed in federally-funded research by the guidelines for the protection of human subjects. (Author)