Fingerprints are extremely useful in personal identification; however, they are usually based on physical rather than chemical images because it remains a challenge to reveal a clear chemical fingerprint easily and sensitively. Herein, a surface plasmon resonance imaging (SPRi) method, combined with a chemically selective stepwise signal amplification (CS3A) strategy, is proposed to chemically image fingerprints with adjustable sensitivity and clarity. High-fidelity glucose-associated fingerprint images were obtained at five to seven cycles of CS3A based on the recognition reaction of concanavalin A (ConA) with dextran. The method is also extendable to image substances that possess and/or can be tagged with ConA- or dextran-recognizable groups. For demonstration, SPRi of carboxylic substances was conducted by amidating the carboxyl group with glucosamine to enable the ConA-based CS3A. Glucose- and carboxyl-based fingerprints were simultaneously and clearly imaged, allowing us to perform quantitative analysis of the representative of either glucose or amino acid (e.g., serine) or both. The curves measured from the standard spots were linear in the ranges of 1-4000 μM for glucose and 3.2-4000 μM for serine, with linear correlated coefficients of 0.9979 and 0.9962, respectively. It was then applied to the study of metabolic secretions in fingerprints during running exercise, yielding variation tendencies similar to those measured from sweat samples in the literature. As a noninvasive tool, the CS3A-coupled SPRi reveals both clear images of fingerprints and quantitative chemical information, and it is anticipated to become a competitive new method for chemically imaging fingerprints.