Background: While modular reconstruction implants can be used to replace the bone lost after bone tumor resection, tumor excision from the neighboring soft tissues can lead to loss of strength and joint range of motion (ROM), which results in worse knee function. Functional recovery after total knee arthroplasty for osteoarthritis has been extensively documented. But few studies have evaluated the recovery after total knee reconstruction following tumor excision despite the fact that most of these patients are young and have high functional demands. We did a prospective cross-sectional study to: 1) compare muscle strength recovery around the knee with an isokinetic dynamometer after tumor excision and reconstruction with a modular implant to the healthy contralateral knee; 2) determine if the differences in peak torque (PT) in the knee extensors and flexors had a clinical impact. Hypothesis: Resection of soft tissues during tumor excision around the knee causes strength loss that cannot be fully recovered. Methods: The 36 patients who underwent extra- or intra-articular resection of a primary or secondary bone tumor in the knee area followed by reconstruction with a rotating hinge knee system between 2009 and 2021 were eligible for this study. The primary outcome was the ability to actively lock the operated knee. The secondary outcomes were the concentric PT during isokinetic testing at slow (90°/sec) and fast (180°/sec) speeds, flexion–extension ROM, Musculoskeletal Tumor Society (MSTS) score, the IKS, Oxford Knee Score (OKS) and KOOS. Results: Nine patients agreed to participate in the study, all of whom had regained the ability to lock their knee postoperatively. PT in flexion and extension on the operated knee was less than the healthy knee. The PT ratio for the operated/healthy knee at 60°/sec and 180°/sec in flexion was 56.3% ± 16.2 [23.2-80.1] and 57.8% ± 12.3 [37.7-77.4], respectively, which corresponded to a slow-speed strength deficit of 43.7% in the knee flexors. The PT ratio for the operated/healthy knee at 60°/sec and 180°/sec in extension was 34.3% ± 24.6 [8.6-76.5] and 43% ± 27.2 [13.1-93.4], respectively, which corresponded to a slow-speed strength deficit of 65.7% in the knee extensors. The mean MSTS was 70% ± 20 [63-86]. The OKS was 29.9/48 ± 11 [15-45], the mean IKS knee was 149.6 ± 36 [80-178] and the mean KOOS was 67.43 ± 18.5 [35-88.7]. Discussion: Despite all patients having the ability to lock out their knee, there was an imbalance in the strength between opposite muscle groups: 43.7% strength deficit at slow speed and 42.2% at fast speed for the hamstring muscles, and 65.7% at slow speed and 57% at fast speed for the quadriceps muscles. This difference is considered pathological with an increased risk of knee injury. Despite this strength deficit, this joint replacement technique, which is free of complications, can preserve good knee function with acceptable knee joint ROM and satisfactory quality of life. Level of evidence: III; prospective cross-sectional case–control study.