Objectives: We devised the Simple postoperative acute kidney injury risk (SPARK) index and classification to predict post-operative acute kidney injury (PO-AKI) in major non-cardiac surgery. The SPARK index has been validated externally in several other centers, but further large-scale validation is needed.  Methods: We performed a retrospective cohort study at two referral centers that were performed the original study. The patients who underwent non-cardiac surgery from 5 departments after the original study period were enrolled. PO-AKI was defined as KDIGO AKI criteria within 7 days after surgery. Critical AKI was defined as KDIGO AKI stage ≥2, post-PO-AKI death, or dialysis within 90 days after surgery. The SPARK estimation was derived from the original model. The SPARK index and score were calculated according to the presented formula. The c-statistic was used to estimate the model function.  Results: A total of 70,579 patients were enrolled. Overall, 3,210 (4.5%) PO-AKI and 656 (0.9%) critical AKI occurred. The prediction power of SPARK index was comparable with SPARK estimation in both AKI (Estimation: area under the curve [AUC] 0.76, 95% confidence interval [CI] 0.747–0.76; Index: AUC 0.75, 95% CI 0.74–0.76) and critical AKI (Estimation: AUC 0.78, 95% CI 0.76–0.80; Index: AUC 0.78, 95% CI 0.76–0.80). The trend was maintained in subgroup analysis by 2 centers and 5 departments. Although SPARK classification was more simplified, its predictive power was decreased. Compared with the original study’s discrimination power, the present study showed compatible prediction power for PO-AKI (AUC 0.75, 95% CI 0.74–0.76) and critical PO-AKI (AUC 0.78, 95% CI 0.76–0.80).  Conclusions: In present large-scale time validation, we validated both the SPARK index and estimation well predicted the occurrence of PO-AKI in our validation cohort. Moreover, the SPARK index was comparable to the SPARK estimation.