Kuei-Yu Chien, Tzu-Han Chan, Ting-Ru Lin, Tzu-Chen Lin
Background: Paired-set (PS) resistance exercise is widely used to increase training density; however, evidence-based guidance remains limited due to heterogeneous exercise sequences and rest configurations. This systematic review aimed to synthesize acute mechanical, training-efficiency, and perceptual responses across distinct PS configurations compared with traditional sets (TSs) in young resistance-trained adults. Methods: Eight databases were searched through 5 August 2026, for acute randomized crossover/parallel trials. PS protocols were categorized into PS-A (alternating exercises with preserved TS rest after each exercise), PS-B (interleaved exercises during recovery), and PS-C (modified post-pair rest). Study-level means and unstandardized percentage differences for volume load (VL), completed repetitions, training efficiency (VL accumulated per unit of time), and rating of perceived exertion (RPE) were synthesized descriptively. Results: Fifteen reports representing 14 studies (223 participants) were included, and all exhibited "some concerns" according to RoB 2. PS-A consistently elicited higher work-output point estimates (VL: 7.9-27.2%; repetitions: 6.3-26.7% vs. TS) with negligible changes in session duration. PS-B consistently reduced session duration by 44.8-50.0%, though repetitions and VL were often compromised depending on the exercise. PS-C responses were configuration-dependent; matched post-pair rest (PS-CMPR) decreased duration (39.7-46.9%) with variable VL (-8.2% to +30.1%) and higher RPE, whereas extended rest (PS-CEPR) showed distinct protocol-specific trade-offs between volume retention and training efficiency. Conclusions: Based on acute trials in young trained populations, PS configurations should be selected by session priority (PS-A for work capacity, PS-B for time constraints, PS-C for versatility). Given the consistent "some concerns" regarding risk of bias, these practical recommendations warrant caution.