Epoxy molding compounds are the most commonly used composite material for encapsulation in the semiconductor industry. They consist of a polymer-based matrix, inorganic filler particles, and a wide range of (in)organic additives to fine-tune the properties. As the encapsulation material is in direct contact with the delicate semiconductor components, the reliability and lifetime of electronic devices heavily depend on applying an appropriate type of encapsulation material. Semiconductor manufacturers typically obtain epoxy molding compounds from specialized external suppliers. Therefore, quality control checking if the correct molding compound was obtained and whether its properties deviate from previous batches is of great interest to the semiconductor industry. In this work, we investigate the capabilities of a Tandem LA-ICP-MS/LIBS approach to comprehensively characterize epoxy molding compounds to construct a classification model capable of distinguishing between different molding compound types. Tandem LA-ICP-MS/LIBS enables detection of the elemental composition of all relevant components within epoxy molding compounds: LIBS can detect polymer-specific signals and the major and minor elements present in the matrix (inorganic filler particles and additives). LA-ICP-MS provides information about elements present at trace levels (e.g., contaminations), which can provide information about batch-to-batch variations. We analyzed 29 samples of 20 molding compound types from 4 different suppliers. Using exploratory data analysis (PCA and HCA), we investigated the spectral fingerprint of the different molding compound types. Finally, a Random Decision Forest-based classifier is optimized and characterized, and a model is constructed. The final classifier is tested with independent samples that were not part of the training set, revealing a satisfying performance and highlighting some molding compound types that are difficult to distinguish. Display omitted •Tandem LA-ICP-MS/LIBS method for quality control in the semiconductor industry.•Exploratory data analysis and development of a classification model.•Machine learning based elemental fingerprinting.