PurposeThe degree to which varus knees can be corrected manually is important when considering total versus unicompartmental knee arthroplasty (UKA). The primary aim was to clarify the relationship between the degree of coronal alignment correction and radiographic parameters involved in UKA prognosis using preoperative full‐length lower extremity valgus stress radiography. The secondary aim was to identify the factors affecting alignment correction.MethodsThis retrospective observational study included 115 knees with medial osteoarthritis that underwent knee osteotomy or arthroplasty. Percent mechanical axis without valgus stress (%MA: neutral, 50%; varus, <50% and valgus, >50%), mechanical lateral distal femoral angle, lateral bowing femoral angle, medial proximal tibial angle (MPTA), joint line convergence angle, medial and lateral joint space width (LJSW) and medial femoral and tibial joint osteophyte size were measured using preoperative full‐length weight‐bearing radiographs. Correlation and multiple linear regression analyses were used to assess associations between parameters and %MA with valgus stress or amount of %MA change (%MA with valgus stress minus %MA without valgus stress).Results%MA with valgus stress was correlated with all radiographic parameters. %MA change was correlated with parameters except for MPTA and LJSW. Multiple regression analyses showed that %MA without valgus stress and MPTA were associated with both %MA with valgus stress and %MA change. When %MA with valgus stress was set at 30%, 40% and 50%, MPTA cutoff values were 81.6°, 83.5° and 84.9°, and cutoffs for %MA without valgus stress were 10.7%, 17.1% and 25.1%, respectively.ConclusionSmall MPTA is strongly associated with less alignment correction under valgus stress in varus knees. The finding is useful in surgical planning, especially to avoid undercorrection with UKA when valgus stress radiographs are unavailable.Level of EvidenceLevel III.