Stature measurements on large samples of unselected 22-year-old Danish males from the period 1852–56 are analysed. Data are given for 18 rural counties and the City of Copenhagen. The distribution of stature deviates significantly and systematically from Normality in all counties. A mixture of two Normal distributions with equal variances describes the empirical distribution of stature in the 18 rural counties and explains important aspects of the distribution of stature in the City of Copenhagen. This mixed structure of the population is accomplished by a factor reducing the mean stature of a fraction of the population by nearly 20 cm. The factor generating this mixed population is probably genetic and inbreeding may play a major part in raising the frequency of affected persons to the observed values, 0·2–2%. The mean stature of the majority of the population and the mixing proportion show significant geographical variation. The frequency of the small subpopulation is generally highest in the least-developed main province, Jutland, and lower on the islands, particularly in and around the City of Copenhagen. On the Zealand group of islands the mean stature of the majority of the population is highest in the most developed area, whereas the reverse occurs in Jutland.