The dataset presents the electron density derived using the Neural-network-based Upper hybrid Resonance Determination (NURD) algorithm (Zhelavskaya et al., 2016) from plasma wave measurements made with the Van Allen Probes Electric and Magnetic Field Instrument Suite and Integrated Science (EMFISIS) (Kletzing et al., 2013). The method employs feedforward neural networks to derive the upper hybrid resonance frequency from the electric field measurements, and hence electron density, in an automated fashion. The dataset contains electron density for the period from October 1, 2012 to January 14, 2018 for RBSP-A and from October 1, 2012 to July 1, 2016 for RBSP-B (RBSP = Radiation Belt Storm Probes). For convenience, the density data are organized in two ways: in terms of orbits and in terms of days. Directories ../../Orbits_organization/ and ../../Days_organization/ contain files with densities per orbit and per day, respectively. Data are provided in .txt and .cdf formats. Data in .mat format are available at ftp://ftp.gfz-potsdam.de/home/rbm/NURD/. For more information on directory organization and files description, please refer to the associated data description and Zhelaskaya et al. (2016).