PURPOSE: The goals of this study were (1) to assess whether the preoperative difference between modalities and extent of deformity are associated with a higher difference between planned and achieved surgical correction and (2) if they yield a higher probability of intraoperative adjustments. METHODS: Retrospective single-centre analysis of patients undergoing patient-specific instrumented (PSI) total knee arthroplasty (TKA). Preoperative radiographic parameters were analysed on weightbearing (WB) long-leg radiographs (LLR) and nonweightbearing (NWB) computed tomography (CT). The 2D/3D difference was calculated as the difference between preoperative WB-LLR (2D) hip-knee-ankle angle (HKA), and NWB CT (3D) HKA. Surgical records were screened to retrieve intraoperative adjustments to the preoperative plan. Postoperative assessment was performed on WB LLR. RESULTS: Two-hundred-eighty-two knees of 263 patients were analysed. The difference of postoperative achieved to planned HKA (HKA CONCLUSION: The extent of deformity is associated with residual coronal deformity following PSI-TKA. Patients with extensive coronal malalignment may benefit from an adaptation of the preoperative surgical plan to avoid unintended postoperative coronal malalignment. Despite the advancements with 3D preoperative planning, intraoperative adjustments in PSI-TKA are frequently performed, in particular in patients with a higher preoperative varus/valgus deformity. LEVEL OF EVIDENCE: Level III.