This paper investigates the performance of indoor terahertz (THz)-aided non-orthogonal multiple access (NOMA) systems in the context of spherical stochastic distances in downlink channel models. With the aim of improving system performance gains, a novel user pairing scheme subject to power control criterion is proposed to adaptively change with regard to the users’ location acquisition and signal’s transmit power. Then, we derive exact closed-form expressions for the user’s outage probability (OP) in order to gain some engineering insights, i.e., key parameters affecting the system performance trend. Monte Carlo simulation is presented to corroborate the accuracy of the theoretical analysis as well as demonstrate the effectiveness of the proposed user-pairing approach.