The new rDNA and DNA-derived "basal" insulin analogs, glargine and detemir, represent significant advancement in the treatment of diabetes compared with conventional NPH insulin. This review describes blood glucose homeostasis by insulin in people without diabetes and outlines the physiological application of exogenous insulin in patients with type 1 and type 2 diabetes. The requirements for optimal basal insulin treatment are discussed and the methods used in the evaluation of basal insulins are presented. An essential criterion in the development of an "ideal" basal insulin preparation is that the molecular modifications made to the human insulin molecule do not compromise safety. It is also necessary to obtain a clear understanding of the pharmacokinetic and pharmacodynamic characteristics of the two currently available basal insulin analogs. When comparing glargine and detemir, the different molar concentration ratios of the two insulin formulations should be considered along with the nonspecificity of assay systems used to determine insulin concentrations. However, euglycemic clamp studies in crossover study design provide a good basis for comparing the pharmacodynamic responses. When the latter is analyzed by results of intervention clinical trials, it is concluded that both glargine and detemir are superior to NPH in type 1 and type 2 diabetes. However, there is sufficient evidence to demonstrate that these two long-acting insulin analogs are different in both their pharmacokinetic and pharmacodynamic profiles. These differences should be taken into consideration when the individual analogs are introduced to provide basal insulin supplementation to optimize blood glucose control in patients with type 1 and type 2 diabetes as well. PubMed-Medline was searched for articles relating to pharmacokinetics and pharmacodynamics of glargine and detemir. Articles retrieved were reviewed and selected for inclusion if (1) the euglycemic clamp method was used with a duration >or=24 h, (2) a single subcutaneous dose of glargine/detemir was used, and (3) area under the curve for insulin concentrations or glucose infusion rates were calculated.