The findings of almost all studies of assortative mating confirm the hypothesis for "like to mate with like" rather than the tendency for "opposites to attract each other". Data in this study were secured from 1000 engaged couples in the Chicago area. All were whites, 3/5 with native born parents. The age range was between 20 and 30. Three fourths of the men and 2/3 of their fiancees were at college level; the rest nearly all high school graduates. About l/2 were Protestants; and a large majority identified themselves with the middle or upper middle class. Two series of personality data were obtained: (1) responses to 42 items in the Thurstone Neurotic Inventory, and (2) Self ratings of 316 of the group on 23 selected personality traits on which the subjects rated themselves along a 5-point scale. Many of the latter traits recurred in spontaneous descriptions of themselves at personal interviews. Data were also obtained on height, wt., health status, complexion, and physical appearance. 14 of the Inventory items were statistically significant in indicating homogamy in regard to neurotic tendencies; a number of other items approached the significance level; and all of the 42 items were in the direction of homogamy. The same was true in the case of the personality ratings: 9 traits were statistically significant; the rest were in the direction of homogamy. Similar trends held for the physical factors studied except in the case of complexion. These same subjects gave evidence of a lesser degree of homogamy in regard to personality characteristics than they did in religious affiliation and behavior, cultural background, courtship behavior, conceptions of marriage, and social participation. "Cultural likeness" appears more important in marital selection than "temper-mental or personality similarity".