The early detection of different diseases is crucial for their successful treatment. Existing literature has established distinct volatile organic compound (VOC) profiles in organic samples that are associated with different disease conditions. These identified profiles hold significant potential for the development of rapid, non-invasive, and cost-effective disease screening tools, meeting the essential criteria for effective diagnostic measures.2-propanol is a secondary alcohol that exhibits different associations with a variety of diseases across diverse organic samples. Notably, elevated levels of 2-propanol in urine samples have been linked to type II diabetes, whereas a decrease in its concentration is observed in cases of malignant biliary strictures. In feces, an increase in 2-propanol is associated with colorectal cancer. Furthermore, disturbances in 2-propanol levels in breath have been correlated with various types of cancers. These findings underscore the potential utility of 2-propanol as a biomarker for disease detection across multiple biological matrices.In this study, we propose a sensor that combines a multimode interference structure with a molecularly imprinted polymer (MIP) functionalization. The multimode sensor is created by fusion splicing a coreless fiber section to a single-mode fiber (SMF). For coating the coreless fiber (CSF), the dip coating technique was utilized, employing a MIP with 2-propanol as the template, and separately, the corresponding non-imprinted polymer (NIP). Our findings reveal that the MIP demonstrates specificity for 2-propanol in the gas phase, showing a sensitivity of 13 pm/wt.% 2-propanol. In contrast, the corresponding NIP lacks sensitivity to 2-propanol. This outcome emphasizes the potential of our proposed sensor design for the selective detection of 2-propanol, thereby highlighting its applicability in gas phase analyses.