BACKGROUND: Salmonella, a notorious foodborne pathogen with a wide range of hosts, poses a significant public health concern globally. Contaminated surface water acts as a potential source of Salmonella transmission. OBJECTIVE: To optimize a Salmonella detection method from large-volume water and analyze surface water samples in Beijing and characterize Salmonella isolates from these samples by whole genome sequencing. METHOD: A microbial enrichment device based on the modified Moore swab (MMS) design was optimized and validated. Thirty five water samples were collected and detected for Salmonella from 11 park lakes, 2 rivers, and 2 farms. Multiple characteristics of isolates were analyzed using antibiotic antimicrobial testing and whole genome sequencing. RESULTS: The optimized MMS unit showed high efficiency (over 80% recovery) and a low detection limit (10° cells) for enriching and isolating Salmonella from large-volume water (10 liters). Compared to the conventional method, the MMS device significantly improved Salmonella detection efficiency (62.86% vs 8.57%) in Beijing's surface water. Most of the Salmonella isolates from surface water belonged to rare serotypes from water creatures and pan-susceptible to the tested antimicrobials. CONCLUSIONS: The study demonstrates the optimized MMS's effectiveness for on-site enrichment of pathogens from large large-volume water, validates the accuracy and sensitivity of Salmonella detection method for surface water, and applies to reveal previously unknown information about Salmonella contamination in Beijing's public water system. HIGHLIGHTS: Salmonella concentrations in water are typically very low, implementation of this method would successfully realize large-volume water sampling and on-site pathogen enrichment, and significantly improve Salmonella detection efficiency in surface water.