Background: Early biomarkers for acute kidney injury (AKI) diagnosis are needed since an increase in serum creatinine levels is a late marker. Neutrophil gelatinase-associated lipocalin (NGAL) is one of the most promising AKI biomarkers. Prior to routine clinical use, it is necessary to evaluate and validate a high-throughput commercially available method for NGAL detection. The aim of this study was to do an independent validation and comparison of the analytical performance of three different commercially available urine NGAL (uNGAL) assays. Methods: Urine samples (n=110) were obtained from various patient groups with and without AKI. All urine samples were processed using Architect NGAL assay, Siemens Advia ® 2400 NGAL test, and Siemens Dimension Vista ® NGAL Test™, based on the three different platforms. Results: Overall, there was good agreement among the three assays: Spearman’s rank correlation coefficient between Architect and Vista was 0.989 (95% confidence interval [CI], 0.983–0.993), between Architect and Advia, 0.962 (95% CI, 0.937–0.977), between Vista and Advia 2400, 0.975 (95% CI, 0.961–0.984). We observed a negative bias of Architect compared with the other assays: comparing Architect to Vista, the mean bias was –55.7 ng/mL (95% CI, –74.3 to –37.0 ng/mL); comparing Architect to Advia 2400, the mean bias was –40.9 ng/mL (95% CI, –56.4 to –25.4 ng/nL). The bias is proportional to the concentration of uNGAL and is more pronounced at higher levels, while irrelevant near the tested cutoff levels of 100 and 190 ng/mL. Comparing Vista and Advia 2400, the mean bias was 10.1 ng/mL (95% CI, 1.5–18.8 ng/mL). Intra-assay imprecision was generally acceptable across all assays; coefficient of variation ranged from 0.8% to 5.3%. Conclusions: All three methods for uNGAL showed acceptable performance for the tested parameters and are comparable with each other at clinically relevant cutoffs. However, Architect yields lower results than the other two methods, with a bias more pronounced at higher uNGAL concentrations, suggesting additional standardization efforts will likely be necessary to better harmonize the uNGAL methods at various clinically relevant cutoffs.