The volatiles in coffee play an important part in the overall flavor profile. In this study, GC–TOF/MS and GC×GC–TOF/MS were used to detect the volatile organic compounds (VOCs) in coffee samples of three different brands at three states (bean, powder, and brew). The differences between the two methods in characterizing VOCs were analyzed using the Venn diagram and PCA (principal component analysis). The important aroma-contributing compounds were further compared and analyzed. The results of the venn diagrams of different coffee samples showed that most VOCs existed in 2–3 kinds of coffee. The PCA of VOCs in different coffee samples showed that the VOCs detected by GC–TOF/MS could distinguish the coffee samples in the different states. GC×GC–TOF/MS was suitable for the further identification and differentiation of the different brands of coffee samples. In addition, pyridine, pyrrole, alcohols, and phenols greatly contributed to distinguishing coffee in three states, and alcohols greatly contributed to distinguishing the three brands of coffee.