To determine and compare the retinal nerve fiber layer (RNFL) thickness, the mean deviation, and the visual field index (VFI) in glaucoma cases with progression detected by spectral domain optical coherence tomography, standard automated perimetry (SAP), and optic disc stereophotographs. The authors studied 246 eyes of 148 patients prospectively (97 glaucoma cases, 132 suspects, and 17 healthy eyes). SAP fields, optical coherence tomography (OCT) images, and optic disc stereophotographs were obtained every 6 to 12 months. Progression was determined in SAP and in OCT with a Glaucoma Progression Analysis software, and also by masked assessment of the stereophotograph series. The Kruskal-Wallis test was applied to evaluate differences between methods in RNFL thickness, visual field (VF) mean deviation, and VFI. The relationship between the baseline classification and the detection of glaucomatous progression by the different tests was assessed by the χ statistic. Ninety-nine eyes (40.2%) presented glaucomatous progression detected by at least 1 examination method. Progressing eyes detected only by OCT had a higher mean RNFL thickness and mean VFI than progressing eyes detected only by VF or stereophotographs (P<0.003). Most progressive cases detected by OCT (68%) were initially classified at baseline as suspects, whereas most eyes with VF progression (61%) were initially classified as glaucoma. The initial classification was significantly related to the presence of progression by different tests [χ (2)=9.643 for VF event analysis and 7.290 for OCT event analysis (P<0.005)]. Different tests are more likely to detect the progression in different clinical circumstances or stages of glaucoma; these should be taken into consideration when performing the difficult task of progression detection.