In order to assess the metabolic efficacy of cyclic nocturnal parenteral nutrition (C‐TPN) in hospitalized patients, a prospective study was carried out with 14 patients having a protein calorie malnutrition due to severe gastrointestinal diseases. Patients population was divided into two groups: one group received 1 milligram per kilogram per day of prednisolone with induced hypercatabolism (n = 4); the second group was not under corticosteroid therapy (n = 10). Patients selected for the study were unaffected by cardiac, hepatic, renal or pulmonary failure, diabetes or bacteriemia. C‐TPN was compared to standard or continuous TPN (S‐TPN) and each type of TPN was 1 week with 24 hours between change over. During C‐TPN, parenteral perfusion was discontinued for 9.1 ± 0.4 hours per day (mean ± SEM). Intravenous regimens were strictly the same during C‐TPN and S‐TPN and provided 192 ± 10 milligram nitrogen per kilogram per day plus mixed energy‐sources equivalent to 50 ± 2 kilocalories per kilogram per day with a glucose‐fat emulsion ratio of 50 ± 5%. Metabolic data for each period included weight, serum albumin and transferrin changes, protein breakdown as estimated by measurement of urea production rate and lean body mass gain estimated from determination of total nitrogen balance. In corticosteroid‐induced hypercatabolic patients, C‐TPN provides visceral protein gain equal to S‐TPN and halves the water‐energy deposits observed during S‐TPN, but is less effective than S‐TPN for decreasing protein breakdown and achieving lean body mass gain. Inversely, in mildly catabolic patients, C‐TPN provides visceral protein and lean body mass gains which are similar to those achieved with S‐TPN; C‐TPN, however, halves the water‐energy deposits when compared to S‐TPN. From these data, C‐TPN can be proposed as an optimal therapy in mildly or uncatabolic patients who do not have severe energy depletion.