Purpose To test a new 2-flash multifocal electroretinogram (mfERG) paradigm in glaucoma using a reduced light intensity of the m-frame flash as opposed to the global flash, as it has been suggested that this may increase the responses induced by the global flash, which has been the part of the mfERG response where most changes have been noted in glaucoma. Methods A mfERG was recorded from one eye of 22 primary open angle glaucoma (POAG) patients [16 normal tension glaucoma (NTG), 6 high tension glaucoma (HTG)] and 20 control subjects. A binary m-sequence (2^13-1, L max 100 cd/m 2 , L min \1 cd/m 2 ), followed by two global flashes (L max 200 cd/m 2 ) at an interval of 26 ms (VERIS 6.0 TM , FMSIII), was used. The stimulus array consisted of 103 hexagons. Retinal signals were amplified (gain = 50 K) and bandpass filtered at 1-300 Hz. For each focal response, the root mean square was calculated. We analyzed 5 larger response averages (central 15°and 4 adjoining quadrants) as well as 8 smaller response averages (central 10°and 7 surrounding response averages of approximately 7°radius each). Three epochs were analyzed: the direct component at 15-45 ms (DC) and the following two components induced by the effects of the preceding focal flash on the response to the global flashes at 45-75 ms (IC-1) and at 75-105 ms (IC-2). Statistical analysis was performed using linear mixed effects models adjusted for age. Results Responses differed significantly between POAG patients and controls in all central response averages. This difference was larger for the central 10°t han for the response average of the central 15°. While these observations held true for all response epochs analyzed, the DC differed least and the IC-1 most when POAG was compared to control. For POAG, the most sensitive differential measure was IC-1 of the central 10°with an area under the ROC curve of 0.78. With a cutoff value of 12.52 nV/deg 2 , 80 % of the POAG patients (100 % HTG, 69 % NTG) were correctly classified as abnormal, while 77 % of the control subjects were correctly classified as normal. When the results of the mfERG were compared to the visual fields, there was a tendency for the mfERG to decrease as the mean defect increased. However, this correlation was only significant in the superior nasal quadrant when the IC-1 of the mfERG was compared to the corresponding area of the visual field. Conclusion When compared to findings from previous studies, reducing the luminance of the m-frame flash in the 2-global flash paradigm did not increase the sensitivity and specificity of the mfERG to detect glaucoma further.