Volatile organic compounds (VOCs) in indoor air are emitted from a wide variety of consumer products and building materials and are routinely inhaled through daily activities. Owing to their effects on human health, identifying VOCs from emission sources associated with indoor air pollution is important for effective exposure management. In particular, VOCs from time-varying emission sources, such as daily activities including cleaning, cooking, and eating, often exhibit complex patterns and require sophisticated methodologies for accurate source identification. This study applied sequential air sampling combined with multivariate analysis to identify VOCs from gradual emission sources, one of the time-varying types, in a workplace. Non-targeted analytical data from sequential indoor air samples were analyzed using principal component analysis (PCA) to visualize indoor air quality patterns. The samples were grouped into three clusters: Group I was characterized by decamethylcyclopentasiloxane (D5), possibly associated with the application of personal care products during the hours when people were present in or near the room; Group II was characterized by toluene, ethyl acetate, and trichloroethylene; Group III was located near the center of the PCA score plot. Notably, individual VOCs found by our method were major components of the temporal changes in total VOC observed during multi-day sampling. This approach can be useful for identifying VOCs from gradual emission sources, particularly those related to occupant activities and product use.