Correlations between preoperative nutritional status on the one hand and postoperative morbidity and mortality on the other hand were reviewed in articles from literature. The inclusion criteria for analysis were the following: large series, well defined nutritional status, major surgery, objective evaluation of postoperative complications and relevant statistical analysis. This study considered factors supposed to alter this correlation, namely the type of primary disease, intensity of denutrition, obesity, site and grade of cancer and age. There was a highly significant correlation between preoperative denutrition and postoperative morbidity and mortality. This correlation was existing whether operation was performed on the gastrointestinal tract or not, whether the primary disease was a cancer or not. The more important the denutrition, the more frequent the postoperative complications and deaths. Morbidity and mortality rates were linearly correlated to denutrition. After head and neck surgery for malignant diseases, malnourished elderly patients had poor prognosis. Obesity had the same prognostic value as denutrition.