Scant attention has been paid in the social sciences to the problem of defining units of analysis. The problem of using culture as a unit of analysis is that culture is not a unit of analysis like a jury is a unit of analysis. It is also a more ambiguous unit of analysis than religion, ethnicity or gender, units which are possible to identify and define. It is concluded that the individual is the least problematic unit for analysis. The limitations of using the individual as the unit of analysis are that group characteristics and behaviors can only be measured indirectly and studies are prone to the ‘individual differences fallacy.’ It is dubious that one can generalize from individuals beyond the community. There are no ultimate primitive units of culture and whatever unit for analysis the researcher selects depends on the questions asked. Always however, a unit of analysis must be clearly defined, it cannot be used as a variable rather variables are extracted from the unit of analysis. Most importantly, there should always be a theory of analysis that justifies the choice of the units for analysis.