We present results of comparisons between consanguineous and control couples and their children in Southern Brasil. Multiple regression, t-test and chi 2 analyses were applied to the data. The following variables were investigated; Parents: structure of first cousin marriages, latitude and longitude of birthplaces, cohabitation time, rurality, marriage age, occupation, number of liveborn children, and frequency of twinning. Schoolchildren: clinical data as classified into 24 discontinuous traits, each one subdivided into two categories (number and severity), and seven continuous traits (weight, height, pulse rate, respiratory rate, systolic pressure, diastolic pressure, and temperature); nutrition indices and biotype; family names; intelligence tests and school performance; birth order, age, hair color, hair type, eye color, and skin color; twinning in the sibship, and infant mortality among sibs of the propositi. Our study showed a) nonrandomness of consanguineous marriages as detected on structural, racial, geographical, and temporal levels; b) a heavy mutational load of about 1.5 lethons acting on infant mortality; c) no inbreeding effect on morbidity in general (with a possible exception of "ear morbidity", including hearing deficit); d) a modest inbreeding depression on height (a decrease of 2 cm with an increase of 10% of inbreeding); 3) a suggestion of inbreeding depression on "intelligence".