Abstract— Nowadays, Fifth Generation-New Radio (5G-NR) is an important research area in telecommunication field. System level studies are essential to evaluate the performance of wireless systems and signaling techniques prior to practical deployment as well as to research purposes. System level studies of wireless systems involve many links between the base station and the user equipment, many modulations and coding schemes and many operation scenarios. Performing system level studies using detailed physical layer simulation using Link Level Simulator (LLS) is impractical and time consuming. Instead, physical layer abstraction models of wireless systems are used in system level studies. Therefore, this paper proposes a physical layer abstraction model for the 5G-NR cellular system following the Effective Signal to Interference plus Noise Ratio (ESINR) approach. The proposed abstraction model uses a novel Exponential ESINR Mapping (EESM) that involves two tunable parameters to model the Physical Downlink Shared Channel (PDSCH) of the 5G-NR system. The proposed model is validated against detailed LLS of the PDSCH. Novel equations are proposed to model the optimum values of the two tunable parameters of the EESM as a function of the RMS delay spread of the wireless channel for different modulation and coding schemes. The obtained results show that the performance – obtained using the proposed EESM abstraction model – is very close to the performance obtained from the detailed LLS. The run time of the abstraction model is independent of the size of the transport block data size. The abstraction model is 10 times faster than the detailed LLS for a transport block size of 3.1 kbits. This factor increases to 153 when the transport block size is increased to 57.376 kbits. The proposed equations for the EESM and its optimum two tunable parameters can be used as an accurate tool to evaluate the Block Error Rate (BER) and the throughput performance of PDSCH in NR wireless system for different environments.