As part of the continuing effort to reassess and improve upon present methods of management of carcinoma of the esophagus and, in particular, to compare the quality and length of survival resulting from several methods of management, both surgical and radiotherapeutic, a study was undertaken of 100 cases selected at random from the files of The Johns Hopkins Hospital from 1951 to 1964 inclusive. Sixteen cases were discarded because of miscoding, lack of a tissue diagnosis, or the presence of non-epidermoid lesions, while in a further 20 follow-up data were judged inadequate. Although retrospective staging was possible in 82 cases, indicating similar extent of disease in the various treatment groupings (Table I), adequate follow-up data were available in only 64. This latter selection formed the definitive study group. Vital Statistics and Clinical Profile of Cases Although the total number of patients in the study group is not large, the data with respect to age, sex frequency, and other historical and physical findings indicate that the series is comparable in all respects to others recently reported (1–4). The average age at diagnosis was 59.7 years with an age range of thirty-eight to eighty-seven, encompassing essentially a normal frequency distribution. Males outnumbered females four to one, and in our hospital population there were approximately equal numbers of Caucasian and Negro patients. Typically, a significant number of patients were either heavy drinkers (25 of 64) or heavy smokers (26), and a majority (44) had in their history one and/or the other finding, both of which have been incriminated as etiologic or associated factors (5). A surprising number of patients (13) were found to have had previous esophageal disease such as hiatus hernia, chronic esophagitis, achalasia, diverticula, and lye stricture. As would be expected, weight loss, dysphagia, and pain were the most frequent presenting complaints and subsequent progression of the disease produced regurgitation in 24 patients prior to initial therapy. Hematemesis and/or melena in 8 cases and hoarseness in 6 became evident later in the pretreatment course of the disease. Analysis of Survival Data Not unexpectedly perhaps, considering the large number of advanced lesions in this series, the crude survival times are generally short (Table II). The majority of patients, regardless of method of treatment, succumbed within twelve months of diagnosis although, as in other series, there were a few long-term survivors (8 of 64 at one year and 2 of 64 at five years) (1–3, 6, 7). The results are presented as median duration of survival after first diagnostic biopsy. Since distributions of survival times for various modalities of treatment were consistently skewed to the low side of the survival scale, it was felt that the median, rather than the mean survival, would be the biologically more significant figure.