TB remains a major global health threat, with diagnostic challenges in low- and middle-income countries. The Fujifilm SILVAMP TB LAM (FujiLAM) assay was developed as a rapid urine-based test for TB diagnosis, but manufacturing inconsistencies in the first generation (v.1) led to the creation of FujiLAM (v.2). This study evaluates the diagnostic performance and consistency of FujiLAM (v.2) compared to FujiLAM (v.1) in Nigerian adults with presumed TB.Methods: We retrospectively analysed 178 urine samples from adults with presumed TB at district hospitals in Abuja, Nigeria. Reference standard used was culture and Xpert MTB/RIF. Sensitivity, specificity, and concordance with the reference standard were calculated for both FujiLAM versions, stratified by HIV status.Results: FujiLAM (v.2) showed an overall sensitivity of 58.3% and specificity of 97.3%, while FujiLAM (v.1) yielded a sensitivity of 52.9% and specificity of 99.0%. These results indicated no significant difference in diagnostic accuracy, confirmed by McNemar’s test (TB-positive samples: χ²=0, p=1.0, 95% CI: 0.013–0.987; TB-negative samples: χ²=0.333, p=0.56, 95% CI: 0.008–0.906).For participants living with HIV, FujiLAM (v.1) detected TB in 5/7 and correctly classified 49/51 TB-negative samples. FujiLAM (v.2) detected 4/7 and 55/56, respectively. For those without HIV coinfection, FujiLAM (v.1) detected 9/17 and 96/97, and FujiLAM (v.2) detected 10/17 TB-positive and 94/97 TB-negative samples.FujiLAM (v.2) showed improved test validity, with fewer invalid results (1 vs. 6). Concordance with the bacteriological reference standard remained high (92.1%, κ=0.623), and overall agreement was 97.1%, showing consistent results between both FujiLAM versions.Conclusions: The second-generation FujiLAM assay showed diagnostic accuracy similar to FujiLAM (v.1). FujiLAM continues to be a promising, non-invasive tool for TB detection in people living with and without HIV. Further large-scale studies are recommended using multiple lots.