Rapid and reliable identification of non-target components in herbal preparations remains a primary challenge, especially when corresponding reference substances are inaccessible. In this work, an efficient post-experiment data processing methodology, named reference substance free diagnostic fragment ion (RSFDFI), was developed based on ultra-high-performance liquid chromatography coupled with linear ion trap-Orbitrap tandem mass spectrometry (UHPLC/LTQ-Orbitrap). The first step of this approach was to cluster the components that share common fragment ions into several groups. After querying the database using a predicted chemical formula, the component with the fewest primary hits was preferentially deduced based on its MS/MS spectrum. Once the structure was characterized, its common fragment ions could be used as the prior structural information to select the possible candidates that would facilitate the subsequent identification for each group. Taking Pudilan Xiaoyan oral liquid (PDL) as a model herbal preparation, which has been extensively used for the treatment of epidemic parotitis and children with hand-foot-mouth diseases, this strategy enables a nearly eight-fold narrowing of the database hits, with fifty-two components, including lignans, flavonoids, alkaloids and steroids, being rapidly identified. In conclusion, our work clearly demonstrates that integrating RSFDFI with high-resolution mass spectrometry is a powerful methodology for rapid identification of non-target components from herbal prescriptions and may open new avenues for chemical analysis in other complex mixtures.