Accurate X-ray structure analysis of two crystals of dehydrated potassium-exchanged A zeolite showed no electron density for the supposed zero-coordinated K atom proposed by Leung et al. Instead, two small electron-density peaks were observed opposite to four oxygen atoms. The peak assigned to 0.5 atom of K(6) at 0.5, 0.24, 0.24 is displaced into the main cage at ∼2.8 Å to two O(3) and ∼3.0 Å to two O(1). The peak assigned to 0.15 atom of K(7) at 0.175, 0, 0 is displaced into the sodalite cage at ∼3.0 Å to four O(3). These distances of 2.8-3.0 Å between electron-density peaks need not apply to interatomic distances because the electron density for the framework oxygens is determined almost entirely by oxygens not bonded to K(6) and K(7). A plausible crystal-chemical argument suggests downward adjustment by ∼0.1-0.2 Å to give interatomic distances. The other potassium atoms are 6.3K(1) at 0.23, 0.23, 0.23; 3K(2) at 0, 0.47, 0.47; and 1.5K(4,5) in an elongated peak centered on 0.16, 0.16, 0.16. All K atoms lie within 3 Å of several framework oxygens, and, although their coordinations are unusual, there is no evidence for zero coordination. An electron microprobe analysis yielded 11.8 Al and 12.2 Si per cell. X-ray diffraction data collected with Cu Ka radiation yielded intensities sufficiently strong for refinement of the superstructure which has regular alternation of Si and Al in the framework. The complex arrangement of K atoms is qualitatively consistent with minimization of electrostatic energy, but may be complicated by weak Si,Al disorder. © 1979 American Chemical Society.