In 1963, a Salmonella Surveillance Program was established jointly by the Centers for Disease Con-trol (CDC), Atlanta, Ga., and the Association of State and Territorial Epidemiologists and Laboratory Directors. As a result, the CDC receives reports from each state every week on the serotypes of Salmonella isolated, the age, sex, and county of residence of the person from whom the organism was isolated, and the source (stool, blood, cerebrospinal fluid) of the isolate. In this paper we review the reports of blood isolates of Salmonella over the past 12 years. For certain serotypes of Salmonella, blood isolates constituted a large proportion of all reported isolates. From 1968 to 1979, 269,704 isolates of Sal-monella from blood and stool were reported to the CDC. These were obtained from 268,507 persons, of which 1,197 persons had both blood and stool isolates reported. Four serotypes accounted for 57.6% of these paired isolates: Salmonella typhimurium, 20.1%; Salmonella typhi, 16.3%; Salmonella heidelberg, 13.6%; and Salmonella enteritidis, 7.6%.

For the 12-year period, 8,018 blood isolates were reported (median, 600 isolates per year; range, 507-1,091 isolates) of 128 serotypes. For 1977, a representative year, 608 blood isolates were reported, or 2.8 isolates per million people (based on US population). The age-specific isolation rates are shown in figure 1. The reported isolation rate from blood was highest for infants younger than one year of age, with the peak rate in the zero-to two-month-old group. For stool iso-lates, the peak rate occurs in infants three to five months of age [1]. For persons from one to 70 years of age, reported blood isolates were uncommon. For persons older than 70 years of age, the reported isolation rate doubled compared with the rate for persons between 60 and 70 years of age, but it was still significantly lower than the rate for persons younger than one year of age. Men ac-counted for 55.8% of the blood isolates but for only 50.7% of the stool isolates, a difference