Two selection experiments were designed to change hematocrit percent. In the first experiment the mice were measured at 60 days and elevated hematocritis were achieved in the first generation with no further response. In the second experiment hematocritis were taken at 100 days and two-way response continued for three generations after which the difference between the high and low lines remained constant. Realized heritability was 71-97 percent in the first generation in one experiment and 24-30 percent for three generations in the second. These compare to an estimate of the degree of genetic determination based on 21 inbred strains of 40 percent. The evidence based on the selection experiments and crosses between inbred strains indicates that few genes are involved in the genetic control of normal hematocrit levels found within inbred strains of mice. The realtionship between hematocrit level and systolic blood pressure suggested by other studies was not a consistent finding in our studies.