The aim of this study was to analyze the quality in the use of total parenteral nutrition (TPN). Retrospective study of patients treated with TPN during one year in a general hospital (n = 92). Assessment of quality was performed through measurements of compliance using 20 criteria. These were analyzed both individually and gathered in 3 groups: adjusted (correct indication), safe (absence of complications) and exact TPN (when prescribed diet matched to the administered one). Criteria compliance was compared to standards values (% expected compliance for a quality service). Patient (age, sex, diabetes mellitus) and process (TPN indication, patient's outcome, type of diet) variables were analyzed (using logistic regression models) as potential factors associated with criteria compliance. Compliance of six criteria was significantly below standard values. Simultaneous compliance of all criteria by group was 17% (95% CI: 10-24) in adjusted TPN, 38% (95% CI: 29-47) in safe TPN, and 94% (95% CI: 90-98) in exact TPN. There was a negative association between an age over 65 years and both a correct indication of TPN (OR= 9.4; 95% CI: 1.1-79.8) and the indicator compliance of all criteria in the adjusted TPN group (OR: 2.9 95% CI: 1.03-8.4). The absence of metabolic complications was influenced (p < 0.05) by the use of standardized diets (OR = 0.3; 95% CI: 0.1-0.9). Quality defects were mainly related to the prescription and the frequency of metabolic complications. In general, TPN is used less appropriately when patients are older than 65 years, and metabolic complications are more frequent when standardized diets are used.