ATP-sensitive potassium (K_ATP) channels are octomeric complexes comprising four subunits each of Kir6.2 and the sulphonylurea receptor (SUR1)[1;2]. Kir6.2 is encoded by KCNJ11 and SUR1 by ABCC8. They control insulin secretion and glucose homeostasis by coupling the metabolic status (ATP/ADP ratio) of the cell to its membrane potential. This ability is conferred by the unique property of K_ATP channels: they are inhibited by ATP and activated by Mg-nucleotides. A rise in blood glucose increases the uptake and metabolism of glucose in pancreatic β-cells, resulting in an increase in the ATP/ADP ratio, which leads to the closure of K_ATP channels, and consequent membrane depolarisation and activation of voltage-gated calcium channels. Entry of Ca²⁺ through the activated calcium channels leads to a rise in intracellular Ca²⁺, which triggers insulin secretion. As glucose levels return to normal, the channels begin to open and reduce insulin secretion. As such, mutations that lead to a loss of function, or impair trafficking to the cell surface, of K_ATP channels lead to congenital hyperinsulinism (CHI), a genetic disorder characterised by the unregulated insulin secretion and severe hypoglycaemia [3]. By contrast, mutations that lead to a gain of function [2], or an increase in the cell surface density [4], of the channels cause permanent neonatal diabetes mellitus (PNDM) where insulin secretion is subnormal and resting blood glucose levels are extremely high. Here we investigated two types of genetic mutation found in Kir6.2: one (E282K) that causes CHI by abolishing surface expression of the channel by preventing the ER exit and the other (Y330C and F333I), found in PNDM patients, increases the cell surface density by inhibiting endocytosis. The E282K mutation prevented the surface expression of K_ATP channels, but did not affect the assembly of Kir6.2 with SUR1. The mutant subunits are retained in the ER, but this retention does not appear to be caused by the mis-folding of the protein. Instead, the mutation abrogates the di-acidic ER exit signal (²⁸⁰DXE²⁸²) in Kir6.2. Using the inhibitory forms of Sar1-GTPases, we demonstrate that wild-type K_ATP channels are recruited into COPII coat vesicles and that this recruitment requires a functional ‘DXE’ ER exit signal [5]. Co-assembly of the mutant subunits with the wild-type Kir6.2 and SUR1 formed functional channels indicating that, in the heterozygous state, the mutant subunits can support the function of the channel, although with diminished ability. These data explain the mild disease phenotype expressed by some members of the family carrying the E282K mutation. The PNDM causing Y330C and F333I mutations abolished the ability of the channel to undergo endocytosis. This defect could be rescued when the wild-type subunits were co-expressed with the mutant subunits. Thus in the heterozygous state, the mutant subunits are able to undergo endocytosis. However, the surface levels of the heterozygous channels were over 2-fold greater than those of wild-type channels, indicating that increased surface channel numbers may contribute to the disease phenotype [4]. In conclusion, our studies with the genetic mutations causing contrasting phenotypic effects illustrate the importance of the regulation of the surface density of K_ATP channels in glucose stimulated insulin secretion.