The molecular structure of trans, planar hydridotrioxygen (HOOO) has been examined by means of isotopic spectroscopy using Fourier transform microwave as well as microwave-millimeter-wave double resonance techniques, and high-level coupled cluster quantum-chemical calculations. Although this weakly bound molecule is readily observed in an electrical discharge of H(2)O and O(2) heavily diluted in an inert buffer gas, we find that HOOO can be produced with somewhat higher abundance using H(2) and O(2) as precursor gases. Using equal mixtures of normal and (18)O(2), it has been possible to detect three new isotopic species, H(18)OOO, HO(18)O(18)O, and H(18)O(18)O(18)O. Detection of these species and not others provides compelling evidence that the dominant route to HOOO formation in our discharge is via the reaction OH + O(2) → HOOO. By combining derived rotational constants with those for normal HOOO and DOOO, it has been possible to determine a fully experimental (r(0)) structure for this radical, in which all of the structural parameters (the three bond lengths and two angles) have been varied. This best-fit structure possesses a longer central O-O bond (1.684 Å), in agreement with earlier work, a markedly shorter O-H bond distance (0.913 Å), and a more acute [angle]HOO angle (92.4°) when compared to equilibrium (r(e)) structures obtained from quantum-chemical calculations. To better understand the origin of these discrepancies, vibrational corrections have been obtained from coupled-cluster calculations. An empirical equilibrium (r(e) (emp)) structure, derived from the experimental rotational constants and theoretical vibrational corrections, gives only somewhat better agreement with the calculated equilibrium structure and large residual inertial defects, suggesting that still higher order vibrational corrections (i.e., γ terms) are needed to properly describe large-amplitude motion in HOOO. Owing to the high abundance of this oxygen-chain radical in our discharge expansion, a very wide spectral survey for other oxygen-bearing species has been undertaken between 6 and 25 GHz. Only about 50% of the observed lines have been assigned to known hydrogen-oxygen molecules or complexes, suggesting that a rich, unexplored oxygen chemistry awaits detection and characterization. Somewhat surprisingly, we find no evidence in our expansion for rotational transitions of cis HOOO or from low-lying vibrationally excited states of trans HOOO under conditions which optimize its ground state lines.