T HE article by Kurtz and Morris1 in this issue of Hypertension raises important issues with regard to the use of genetically hypertensive rats. These issues are 1) the origins of various stocks, 2) the definition and maintenance of the integrity of various stocks of interest in hypertension research, and 3) the construction, appropriateness, and use of control strains for comparison with hypertensive strains. The article by Kurtz and Morris 1 traces the tortured origin of the Wistar-Kyoto (WKY) "strain" as a control for spontaneously hypertensive rats (SHR); clearly, the need for a control strain here was an afterthought and not all WKY are likely to be genetically identical. Their article illustrates the frustration of leaving unanswered the question of how to obtain genetic information from a hypertensive strain.The problems relating to control strains originate, in my view, from the unrealistic and inappropriate expectations imposed on a comparison of a hypertensive and control strain rather than from the undefined origins and possible genetic inappropriateness of controls. It is obvious from the literature on SHR that many investigators compare SHR and WKY for some biochemical or physiological trait (call it "trait X") and then conclude that the strain difference in trait X may be causally related to blood pressure differences. The problem is that, with this kind of information, the strain difference in trait X may be the result (rather than the cause) of strain differences in blood pressure, or strain differences in trait X may be due to genetic drift. Genetic drift is the chance selection and fixation of the contrasting genes controlling trait X in the two different strains. Strain differences arising from genetic drift have nothing to do with strain differences in blood pressure.If trait X differs between a hypertensive and a normotensive control strain the appropriate conclusion is simply that there are strain differences in trait X and there are strain differences in blood pressure. Why make such comparisons if this is all that can be concluded? Obviously, one has to determine if strain differences exist in a trait of interest as a first step in determining if there is any difference worthy of further study. If there is no strain difference in trait X, then there is no genetic diversity to study. To unravel the cause and effect relationship between trait X and blood pressure, however, one must be willing to do more than make strain comparisons.The meaning of strain-differences in trait X vis-a-vis strain differences in blood pressure can be found by determining if trait X and blood pressure are genetically separable or inseparable. That is, one tests in breeding experiments whether the association of trait X and blood pressure (i.e., a component of blood pressure) is retained in genetically segregating populations (i.e., populations in which genes can recombine at random). If the gene (or genes) controlling trait X is one of the genes influencing blood pressure, then it will remain associated with an i...