The neutron-capture reaction is of fundamental use in identifying and analyzing the γ-ray spectrum from an unknown object as it gives unambiguous information on exactly what isotopes are absorbing the neutrons. There are many applications where this can be used passively (nonproliferation), or actively where an external neutron source is used to probe an unknown assembly (planetary studies). There are known capture-γ data gaps in the ENDF libraries used by transport codes for various nuclear applications. A new database, EGAF, containing thermal neutron-capture γ-ray data is used to improve the capture-7 information in the ENDF libraries. For many nuclei the unresolved quasi-continuum part of the γ cascade is not available experimentally. In this work, we have modeled this contribution using the Monte Carlo statistical-decay code DICEBOX, in addition to improving level-scheme evaluations. For capture of higher-energy neutrons there is little experimental data available, making evaluation of modeling codes problematic. We plan to continue the DICEBOX approach through the resolved resonance region where spin and parity information is partially known. In the unresolved resonance region, and up to 20-MeV incident neutron energy, we are applying Hauser-Feshbach models to predict the capture-γ spectrum.