The authors have studied the electronic structure of InN and GaN employing G 0 W 0 calculations based on exact-exchange density-functional theory. For InN their approach predicts a gap of 0.7 eV. Taking the Burnstein-Moss effect into account, the increase of the apparent quasiparticle gap with increasing electron concentration is in good agreement with the observed blueshift of the experimental optical absorption edge. Moreover, the concentration dependence of the effective mass, which results from the nonparabolicity of the conduction band, agrees well with recent experimental findings. Based on the quasiparticle band structure the parameter set for a 4 ϫ 4 k · p Hamiltonian has been derived. © 2006 American Institute of Physics. ͓DOI: 10.1063/1.2364469͔The group III-nitrides AlN, GaN, and InN and their alloys have become an important class of semiconductor materials, in particular, for use in optoelectronic devices such as green and blue light emitting diodes and lasers. Among the three materials InN is still the least explored, due to difficulties in synthesizing high quality single crystals. Only very recently these problems have been overcome, 1 but many of the key band parameters have not been conclusively determined until now.1,2 The most controversially discussed parameter is currently still the fundamental band gap of InN. For many years it was believed to be approximately 1.9 eV, but essentially any value between 0.65 and 2.3 eV has been reported in the literature over the last 30 years.1 However, more recent experiments on high quality samples grown by molecular beam epitaxy and recent ab initio calculations support a significantly lower value around 0.7 eV. [3][4][5][6][7] Different hypotheses have been proposed to explain the large variation in the measured band gaps. Defects could be responsible for inducing states in the band gap or give rise to a pronounced Burnstein-Moss effect due to a shift in the Fermi level caused by a high intrinsic electron density. Nonstoichiometry may increase the defect concentration or alter the crystaline structure. The formation of oxides and oxynitrides would increase the band gap, whereas the precipitation of In clusters leads to additional features in optical absorption spectra. 1,8 In this letter we demonstrate that first principles calculations can contribute to the solution of this fundamental question. By combining density-functional theory ͑DFT͒ with many-body perturbation theory in the G 0 W 0 approximation, 9 which is currently the method of choice for calculating quasiparticle excitations in solids, 10,11 we combine atomistic control over the material with accurate calculations for the band structure and the band gap of stoichiometric and defect-free structures.Previous ab initio studies were aggravated by the fact that DFT calculations in the local-density approximation ͑LDA͒ predict InN to be metallic in the zinc blende and wurtzite structures. Subsequent G 0 W 0 calculations only open the gap to 0.02-0.05 eV, 12,13 while adding self-interaction correcti...