Purpose <p>To analyze the correlation and degree of agreement between the patellar heights measured at knee flexion of 0° and 30° before and after total knee arthroplasties (TKA) without patellar resurfacing.</p> Methods <p>One hundred primary TKAs with nonresurfaced patella were prospectively evaluated. We measured the Insall–Salvati ratio (ISR), modified ISR (mISR), Blackburne-Peel ratio (BPR), and Caton-Deschamps ratio (CDR) in weightbearing true lateral radiograph at knee flexion of 0° and 30°. The correlations between the patellar height measured at knee flexion of 0° and 30° were analyzed by Pearson correlation analysis. The degree of agreement between the patellar heights in knee flexion angles of 0° and 30° was analyzed using the limits of agreement (LoA) of Bland–Altman analysis; a difference of &lt;  ± 0.2 between the measurements at knee flexion of 0° and 30° was deemed clinically acceptable.</p> Results <p>Very strong correlations existed between the pre- and postoperative ISR (r = 0.826 and 0.823, <i>p</i> &lt; 0.001, respectively), and the preoperative mISR (<i>r</i> = 0.802, <i>p</i> &lt; 0.001) measured between knee flexion angles of 0° and 30°. Strong correlation was observed in the other pre- and postoperative measurements. The range between upper and lower LoAs of the ISR measured at knee flexion of 0° and 30° was &lt;  ± 0.20 preoperatively (−0.174 ~ 0.155) and postoperatively (−0.181 ~ 0.174), while the ranges for all other measurements did not lie within the clinically acceptable range.</p> Conclusions <p>The ISR measured at 0° or 30° of knee flexion during weightbearing can be interchanged reasonably before and after TKA. Although knee flexion of 30° is the standard, lateral radiograph of 0° flexion can be a reasonable alternative for evaluating patellar height using the ISR.</p> <p><i>Level of Evidence:</i> Level II.</p>