The selective molecular recognition of chloride versus similar anions is a continuous challenge in supramolecular chemistry. We have designed and prepared a simple pseudopeptidic cage (1 a) that defines a cavity suitable for the tight encapsulation of chloride. The interaction of the protonated form of 1 a with different inorganic anions was studied in solution by (1)H NMR spectroscopy and ESI-MS, and in the solid state by X-ray diffraction. The solution binding data showed that the association constants of 1 a to chloride are more than two orders of magnitude higher than to any other tested inorganic anion. Remarkably, 1 a displayed a high selectivity for chloride over other closely related halides such as bromide (selectivity=111), iodide (selectivity=719), and fluoride (selectivity >1000). Binding experiments ((1)H NMR spectroscopy and ESI-MS) suggested that 1 a has a high-affinity (inner) binding site and an additional low-affinity (external) binding site. The supramolecular complexes with F(-), Cl(-), and Br(-) have been also characterized by the X-ray diffraction of the corresponding [1 a⋅nHX] crystalline salts. The structural data show that the chloride anion is tightly encapsulated within the host, in a binding site defined by a very symmetric array of electrostatic H-bonds. For the fluoride salt, the size of the cage cavity is too large and is occupied by a water molecule, which fits inside the cage efficiently competing with F(-). In the case of the bigger bromide, the mismatch of the anion inside the cage caused a geometrical distortion of the host and thus a large energetic penalty for the interaction. This minimalistic pseudopeptidic host represents a unique example of the construction of a simple well-defined binding pocket that allows the highly selective molecular recognition of a challenging substrate.