Our knowledge of interracial marriage in the United States is fragmentary, inadequate and fraught with contradictions. A major methodological finding of this study, discovered by a comparison of statistical records for Philadelphia (1960-1962 and 1965-1966) with marriage license applications, is that there has been a 32 percent error in reporting mixed race cases. The full significance of this as regards existing data can only be conjectured at present. In Pennsylvania, it would seem, areas of high concentration of nonwhites show the lowest intermarriage rates. In the state, excluding Philadelphia, about 3 out of 4 mixed marriages involve nonwhite males; in Philadelphia, the figure is 52 percent. To some extent nonresidents seem to be attracted to Philadelphia for their intermarriages; but, on the other hand, a considerable number of the 84 percent who are residents declare to having the same address. As measured by the interval from application to performance of the ceremony, they do not marry in haste, nor do they show a strong urge to use their license elsewhere in the state. There is no remarkable age disparity for these couples. They do marry somewhat later—about 2 years later for those who are entering upon their first (primary) marriage; and a large proportion of the couples show a prior divorce experience. These data for Philadelphia and Pennsylvania disclose a tendency, noticed also in other studies, for the rate of such marriages to increase, so that now about 2 percent of Philadelphia and nearly 5 percent of Pennsylvania nonwhites are marrying interracially.