In this study, chlorophyll fluorescence (ChlF) was used to detect the onset of soybean plant injury from treatment of glyphosate, the most widely used herbicide. Thirty-six pots of nonglyphosate-resistant soybean were randomly divided into three groups and treated with different doses of glyphosate solutions. The three treatment groups were control (CTRL) group (with no glyphosate treatment), 0.25X group (treated with 0.217 kg·ae/ha solution of glyphosate), and 0.5X group (treated with 0.433 kg·ae/ha solution of glyphosate). Three kinds of fluorescence measurements, steady-state fluorescence spectra, Kautsky effect parameters, and ChlF-related spectral indices were extracted and generated from the measurements in the glyphosate treatment experiment. The mean values of these fluorescence measurements for each of the CTRL group, the 0.25X group, and the 0.5X group were calculated. Glyphosate-induced leaf injury was then analyzed by examining the separability of these mean values at 6, 24, 48, and 72 hours after the treatment (HAT). Results indicate that the peak position of far-red ChlF shows an obvious blue shift for glyphosate-treated soybean, and peak values of steady-state fluorescence spectra for the three groups can be significantly distinguished from each other at 48 HAT and later. Four Kautsky effect parameters, Fv, Fv/Fm, Area, and PI, are parameters sensitive to glyphosate treatment, showing some differences between the CTRL group and treated groups at 24 HAT, and significant differences among the three groups at and beyond 48 HAT. Moreover, ChlF-related spectral indices, R2683/(R675·R690) and R690/R655, are also shown to be useful in detection of the glyphosate injury, though they are less effective than the steady-state fluorescence spectra and the Kautsky effect parameters. Based on the presented results, it can be concluded that glyphosate-induced soybean injury can be detected in a timely manner by the ChlF measurements, and this method has the potential to be further developed into practical use.