The cancer risk changes of 19 human neoplasias over time and space, as expressed in terms of the logarithm of age-adjusted incidence rate (log AAIR), were found to hold a linear correlation with each other--a finding suggesting that the distribution pattern of log AAIR data sets of 2 cancers, when plotted on a two dimension diagram, may show a good fitness to the chemical equilibrium model a product of the law of mass action. On the basis of the statistical analysis of the data, we reached the conclusion that the risk changes of a given neoplasia in space represents the function of the centripetal force of an activated oncogene and the centrifugal force of an inactivated tumor suppressor gene, both of which should cooperate with each other to create a thermodynamic equilibrium under the law of mass action. The purpose of this study was to test the contribution of the oncogene-tumor suppressor gene complex to the sex discrimination of cancer risk in 19 human neoplasias. The results obtained are as follows: a) the correlation coefficient r seq of the sequential regression analysis, as applied to 47 log AAIR data sets of one tumor pair, served as a criterion in testing the balance of power between oncogene activation and tumor suppressor gene inactivation. Sole activation of the oncogene should give an r seq value of -1.0, whereas sole inactivation of the tumor suppressor gene should give an r seq value of +1.0. b) Esophageal cancer and laryngeal cancer, two sex-discriminating tumors with distinct male predominance were each associated with differential implications of the oncogene-tumor suppressor gene complex between the male and female populations: in both tumors, the male populations were associated with a complex of activated oncogene and inactivated tumor suppressor gene, whereas the female population was associated with another complex of weakly activated (esophageal cancer) or non-activated (laryngeal cancer) oncogene and inactivated tumor suppressor gene, as assessed by the r seq criteria. c) The intersex correlation of cancer risk in both esophageal cancer and laryngeal cancer for 47 populations throughout the world, was rather weak, when compared with other members of human neoplasias. The intersex difference of r seq as expressed in terms of t value of Student's t test for each of 19 human neoplasias, was negatively correlated with the correlation coefficient r of the intersex regression analysis with the same 47 populations. It was indicated that a change in the intersex linkage of cancer risk may be related to the differential implication of the oncogene-tumor suppressor gene complex in carcinogenesis. In summary, we conclude that the hormonal milieu of the host plays a cardinal role as the modifier of the oncogene-tumor suppressor gene impact.