Twenty-three patients between 31 and 76 years of age suffering from either hypertension or glaucoma underwent automatic perimetry with Octopus programs 31 or 32, followed by program G-1. This was done to compare the information content of the latter, new program with that of the former two, long-used in different combinations. Only one eye of any one patient was considered in this study. Using an evaluation program, the G-1 program calculates the so-called "field indices" which help to interpret the results. These field indices are the mean differential light sensitivity threshold (mDLS), the mean defect (MD), the loss variance (LV), and the corrected loss variance (corrected for short-term fluctuations (CLV)). In order to compare programs 31, 32, and their combination 31 + 32, with the field indices of program G-1, the mean DLS was taken from the Delta evaluation program, while the mean defect, the loss variance, and the corrected loss variance were calculated on the basis of programs 31, 32, and the combination of them. The mDLS and mean defect determined with program G-1 were significantly different from those found with programs 31, 32, and the combination of them. On the one hand, this was due to the difference in test patterns, that of G-1 being more centrally weighted; on the other hand, this is indicative of the precision and discrimination capability of Octopus measurements. Calculation of loss variance and corrected loss variance reveals no significant difference between the examination methods as regards these field indices, that is, between G-1 and 31, 32, or 31 + 32.(ABSTRACT TRUNCATED AT 250 WORDS)