The KCoEM (Korean Color–Emotion Mapping) dataset is a structured color–emotion mapping resource that aligns the 44 emotion categories of KOTE with their corresponding Korean and English emotion labels. Rather than arbitrarily assigning colors to emotion categories, the dataset was systematically constructed through an extensive review and integration of empirical findings from prior studies in color psychology and emotion–color association research. The resulting dataset consists of a total of 235 emotion–color mapping entries, providing a grounded reference for linking linguistic representations of emotion with their visual color attributes.