Purpose To compare the strength of the structure–function association between scanning laser polarimetry-measured retinal nerve fiber layer (RNFL) thickness and visual sensitivity. Two methods of corneal compensation were evaluated, variable corneal compensation (VCC) and enhanced corneal compensation (ECC). Design Observational case series. Participants One hundred twenty-seven glaucoma (repeatable abnormal visual fields [VF] by pattern standard deviation and/or glaucoma hemifield test) or glaucoma suspect (glaucomatous-appearing discs by photograph assessment without field defects) participants in the University of California, San Diego Diagnostic Innovations in Glaucoma Study. Methods One eye of each participant was imaged using GDx VCC and GDx ECC on the same day. Visual fields tested using the Humphrey Field Analyzer (with Swedish interactive threshold algorithm) were obtained within 6 months of imaging. Main Outcome Measure The associations ( R 2) using linear and logarithmic regression between RNFL thicknesses measured in 6 sectors (inferior, inferotemporal, temporal, superotemporal, superior, nasal) with VCC and ECC and VF sensitivities (decibel threshold measurements) measured in 6 corresponding sectors were compared. Comparisons were made using paired t tests on the log-transformed absolute values of regression residuals. Results Using GDx VCC, 32 scans had a typical scan score (TSS) ≤ 78 (lowest quartile) and no ECC scans had TSS<86. Most RNFL thickness measurements obtained using GDx VCC were significantly greater than those measured using GDx ECC. Regional structure–function associations ( R 2) ranged from 0.03 (temporal RNFL) to 0.22 (superotemporal RNFL) for VCC and from 0.01 (temporal RNFL) to 0.26 (superotemporal RNFL) for ECC. Associations generally were slightly stronger for ECC than for VCC, although these differences were only significant for inferotemporal RNFL ( R 2 = 0.19 and 0.11, for ECC and VCC, respectively). Conclusion The RNFL thickness associations with VF sensitivity are stronger using ECC compared with VCC, suggesting that ECC provides a better cross-sectional representation of visual function than VCC.