Living organisms display a spectrum of wondrous colors, which can be produced by either pigmentation or by structural coloration. A relatively well-studied system, which produces colors via an array of alternating anhydrous guanine crystals and cytoplasm, is responsible for the metallic luster of many fish. The structure of biogenic anhydrous guanine was so far believed to be the same as that of the synthetic one - a monoclinic polymorph (denoted as α). Here we re-examine the structure of biogenic guanine, using detailed experimental X-ray and electron diffraction data, exposing troublesome inconsistencies - namely, a ‘guanigma’. To address this, we sought alternative candidate polymorphs using symmetry and packing considerations, then utilized first principles calculations to determine whether the selected candidates could be energetically stable. We identified theoretically a different monoclinic polymorph (denoted as β), were able to synthesize it, and to confirm using X-ray diffraction that it is this polymorph that occurs in biogenic samples. However, the electron diffraction data was still not consistent with this polymorph, but rather with a theoretically generated orthorhombic polymorph (denoted as γ). This apparent inconsistency was resolved by assuming that the electron diffraction pattern is dominated by crystal stacking faults composed of two offset molecular layers.