Bivalve molluscan shellfish (BMS) contamination with gastroenteritis viruses such as norovirus (NoV) is recognized as a significant public health risk worldwide. These foodborne epidemics are characterized by very low viral concentrations in the implicated foods, and by diverse viruses or multiple NoV strains originating from human sewage, resulting in different strains (co-)infecting the consumers. Next-generation sequencing (NGS) offers promising means to describe the diversity of strains present in BMS or to retrace transmission chains in outbreak settings, but their sensitivity and reproducibility remained to be assessed for this application. In this work, we evaluated the ability of three promising NGS methods to sequence diverse NoV in environmental or food BMS and human stool samples. Using laboratory-prepared samples of known NoV composition, we evaluated the sensitivity, reproducibility, repeatability and selectivity of metabarcoding, capture-based metagenomics and long amplicon sequencing, considering representative NoV strains from genogroup I and II, and the impact of the BMS matrix. The metabarcoding, with separate amplification of polymerase and capsid gene segments followed by Illumina sequencing, was the most sensitive method. It was applied to a selection of 212 BMS samples collected through the European Commission’s NoV baseline survey (BLS), demonstrating a high diversity of NoV sequences found in the BMS which reflect the diversity of NoV strains circulating in the European human population. Besides, a capture-based metagenomics with enrichment of vertebrate viruses was applied on 20 of these BLS samples as well as 20 BMS linked to outbreaks and 10 related human stool samples. In BMS, it yielded NoV sequences compatible with the genomes identified in stool samples, but they were too short to allow definitive confirmation of the infection source. The present report describes NGS methods, including the bioinformatic pipelines, applicable to molecular epidemiology of NoV in BMS, their current limitations and expected outcomes