Gluten monitoring is essential to protect consumers requiring gluten avoidance, but interpretation of multicentre surveillance data may be influenced by changes in sampling intensity, food categories and monitoring contexts. This study characterised gluten-monitoring activities conducted by nine Italian institutions between 2021 and 2025 and evaluated temporal and food-category patterns of non-compliance, complemented by proficiency-testing (PT) data from the CReNaRiA laboratory network. Aggregated monitoring data were analysed by year, food category and sampling context, and samples containing >20 mg/kg gluten were classified as non-compliant. Overall, 10,376 samples were analysed, of which 94 (0.91%) were non-compliant. Pasta, bakery and pastry products, and ready-to-eat foods accounted for 84.0% of non-compliant samples. Annual non-compliance remained below 1% between 2021 and 2023 and was higher in 2024 and 2025. Temporal variation was supported overall (p = 0.005), but these years showed different patterns: the 2024 signal remained evident after sensitivity analysis, whereas the 2025 increase was largely attributable to a concentrated bakery and pastry monitoring context. CReNaRiA PT exercises showed consistently high interlaboratory qualitative agreement across the study period. Overall, gluten non-compliance was infrequent but unevenly distributed across food categories and over time. Integrating analytical results with harmonised information on sampling context, claim status, product characteristics and traceability may strengthen targeted, risk-informed gluten surveillance.