Purpose The value of computer navigation in total knee arthroplasty (TKA) for arthritic knees continues to be debated. The purpose of this study was to evaluate the value of navigated TKA associated with updated alignment philosophy. Methods This prospective randomized controlled trial enrolled 38 consecutive patients (76 knees) and were randomly assigned to both groups. The demographic data and perioperative data were recorded. The coronal plane alignment of the knee (CPAK) classification was used to classify knee alignment phenotypes. Radiographic outcomes were measured and subgroup analysis was further performed. Clinical outcomes were evaluated using patient-reported outcome measures (PROMs). Surgery-related complications were recorded. Results The distribution of CPAK phenotypes following constitutional aligned TKA was equivalent to the native cohort, whereas the mechanical aligned TKA dramatically altered the phenotype distribution from type I and type II to type V and type IV. Final implant positioning was different between groups, with constitutional aligned TKA having larger cTCA ( P  = .004), joint line obliquity ( P  = .006), joint line distance ( P  = .033) and smaller sFCA ( P  = .013). Subgroup analysis showed higher actual accuracy of component positioning was achieved in navigated TKA, especially in knees with deformity of > 10° ( P  < .05). Patients reported higher HSS score at three months postoperatively in constitutional aligned group ( P  = .002). One patient in navigated group suffered femoral pin site fracture caused by a minor trauma. Conclusion Computer navigated TKA allows for restoration of constitutional alignment and minimizes soft tissue release, which when compared to mechanical alignment may be associated with superior early outcomes.