Purpose: Analyze the stability and accuracy of 3 perimetric techniques. Methods: A total of 104 stable eyes (65 subjects) with ocular hypertension and early glaucoma [group G, mean defect = 1.08 dB, SD = 2.0, in standard TOP automatic perimetry (SAP)] were examined 5 times during 18 months using: (a) SAP; (b) Pulsar temporal modulation perimetry (T30W), and (c) frequency-doubling technology (FDT N30). Ninety eyes from 90 normal controls were compared with the first set of examinations of group G. Results: The learning effect was minimal in the 3 techniques but higher in Pulsar (1.0 src, p < 0.05) than in SAP and FDT (0.4 dB). Long-term fluctuation (F) was significantly higher in FDT (3.1 dB, SD = 1.4, p < 0.0001) than in SAP (2.3 dB, SD = 1.1) and in Pulsar (1.9 src, SD = 0.7). Pulsar and FDT reduce F when increasing the number of examinations. F seems equivalent in SAP and FDT and lower in Pulsar, considering small-scale differences of the 3 perimeters. Conclusions: A slight learning effect would be expected on FDT and SAP in patients with previous experience with SAP. The stability and sensitivity of Pulsar is greater than on the other 2 systems. For early diagnosis of glaucoma it is essential to prove the reproducibility and coincidence of perimetric results.