Minimally invasive methodology, mathematical model, and software for analysis of glucose homeostasis by deconvolution of insulin secretion, hepatic extraction, post-hepatic delivery, and sensitivity from 24-hour standardized meals test have been developed and illustrated by the study of glucose homeostasis of a genetically based leptin-deficient patient before and after leptin replacement treatment. The only genetically leptin-deficient adult man identified in the world was treated for 24 months with recombinant methionyl human leptin. Blood was collected every 7 minutes for 24 hours, with standardized meals consumed during the 4 visits: at baseline, one-week, 18-months, and 24-months after initiation of the treatment. Concentrations of insulin, C-peptide, and plasma glucose were measured. Insulin secretion was obtained by deconvolution of C-peptide data. Hepatic insulin extraction was determined based on our modifications of the insulin kinetics model . Insulin sensitivity for each of the four meals was calculated by using the minimal glucose model approach. Hepatic extraction of insulin was the first element of glucose homeostasis to respond to leptin replacement treatment and increased 2-fold after one week of treatment. Insulin secretion and delivery rates decreased more than 2-fold and insulin sensitivity increased 10-fold after 24 months of treatment. Computer programs for analysis of 24-hour insulin secretion, extraction, delivery, and action are available upon request.