Covert measures of the family system can tap concepts and discover patterns of which family members themselves are unaware and can avoid social desirability bias. Covert measures can be particularly useful in cross-cultural research and can avoid many pitfalls of language and translation. Three such measures which have been used cross-culturally are described. The first is a method of measuring triangulation of children by the parents based on patterns of agreement in a true-false questionnaire describing the family. The second is a projective measure in which family members as a group create a picture of their family distinguishing relationships between individuals and boundaries within and around the family. The third is a global coding scheme used to evaluate family processes from taped interaction.Key words: family assessment; covert measures; cross-cultural comparisons Families are complex social systems. Many concepts that describe individuals, like height, gender, race, even happiness, can be reasonably easily seen and measured. Family systems concepts, on the other hand, are less directly observed and measured. Some of these family concepts may be recognized within the culture. When the culture gives prominence to a concept, people within that culture can readily recognize and report on this concept (Baldwin, 1992), like the way U.S. families are attuned to individuality and independence, or Japanese families are attuned to interdependency and context. Other concepts are more difficult for family members to see--and for researchers to measure--when the culture does not call attention to them, like boundaries or a concept like positive dependency (Doi, 1981). Creative, careful overt measurement can still be possible, but when transparent, overt measures cannot be relied upon or when subjects are unable or unwilling to report honestly or completely, measures which are less transparent, or covert, can be of value.In this paper we describe some covert measures developed during a 40-year cross-cultural research project based on structured whole-family interviews conducted in family homes with families of adolescents in the U.S. and Japan. We first review issues involving the measurement of family concepts and suggest how covert measures low in transparency can ameliorate some concerns. We then describe the family systems theory informing this research and give three examples of different kinds of covert measures we have used, one based on patterns of agreement on a