Objective This study was done to determine the validity of amsler grid test black on white (BOW), as well as white on black (WOB) for identifying central visual field (VF) defects in patients with advanced glaucoma. Design Prospective study. Participants We prospectively included 100 consecutive eyes of 88 adult patients with advanced glaucoma and 100 eyes of 100 normal individuals. We used a lottery method to choose the side of the eye for the control groups. Methods All participants had reliable Humphrey 10–2 Swedish Interactive Threshold Algorithm (SITA) standard VF. Both the BOW and WOB amsler grid tests were done for each group. Sensitivity, specificity, and positive and negative predictive values of the amsler grid scotoma area were calculated with the 10–2 VF as the reference standard. Results The mean ± standard deviation (SD) of age and the 10–2 VF mean deviation (MD) of advanced glaucoma eyes were 59.8 ± 11.8 (range 34–84) years and -19.94 ± 9.8(range -34.98–-0.52) respectively. Among 108 eyes with normal 10–2 VF test, 103 had a normal BOW amsler grid test and 5 had an abnormal BOW test. Among 92 eyes with an abnormal 10–2 VF test, 74 had an abnormal and 18 had normal BOW amsler grid test. Sensitivity, specificity, and positive and negative predictive values of the BOW amsler grid test were 80.4%, 95.4%, 93% and 85.1% respectively whereas that of the WOB amsler grid test were 71.7%, 95.4%, 93% and 72.8% respectively. Conclusion The sensitivity and specificity of both BOW and WOB amsler grid tests were high in detecting VF defects in advanced glaucoma.