This study aimed to examine quarter-to-quarter variations in external and internal loads, the relationships between RPE and external loads, and the effects of contextual factors on external and internal loads in backcourt and frontcourt players. 16 professional male basketball players (8 backcourt, 8 frontcourt) from the Chinese National Basketball League were recruited. External load was quantified using Catapult S7 devices to record PlayerLoadTM (PL) and Inertial Movement Analysis (IMA) metrics, while internal load was assessed via rating of perceived exertion (RPE). From the first to the fourth quarter, frontcourt players displayed significant reductions in change of direction to the right (Q1: 0.48 ± 0.29 vs. Q4: 0.33 ± 0.16; p < 0.05) and explosive efforts (Q1: 1.49 ± 0.40 vs. Q4: 1.04 ± 0.35; p < 0.01). In contrast, backcourt players reported higher RPE (Q1: 4.28 ± 1.04 vs. Q4: 5.13 ± 1.05; p < 0.05) and recorded higher PL (Q1: 103.10 ± 44.59 vs. Q4: 123.21 ± 57.66; p < 0.01) in the fourth quarter. Frontcourt players showed higher RPE, PL, and PL · min−1, especially in the second quarter, but lower jump-related (IMA jump low, medium, total) and acceleration metrics than backcourt players. RPE correlated moderately to largely with PL and high-intensity jumps (r = 0.41–0.72, all p < 0.05). Backcourt players reported higher RPE in wins compared with losses (ES = 0.42; p < 0.05). These findings reveal notable positional differences in game load profiles, underscoring the need for context specific load management to optimize position-specific recovery.