An Arabic version of the 70-item Personal Friends Questionnaire was given to 1028 Egyptians. Each item, on the characteristics of personal friends, had to be given one of 5 ratings ranging from "very necessary" (+2) to "definitely opposed" (-2). The extreme response set scores were suggested as possible measures of intolerance of ambiguity, and a general hypothesis formulated according to which social groups with higher tension levels would earn higher extreme response scores than social groups with lower tension levels. Predictions were made and tested about the effects on extreme response set scores of age, sex, religious affiliation, and socioeconomic status. In general, the main hypothesis was confirmed.