Multi-year series of surface microplastics at fixed stations are scarce in the Mediterranean, so episodic variability cannot be separated from persistent contamination. This study aimed to quantify the interannual variability of surface microplastics in Sicilian coastal waters, test differences among sites and along the inshore–offshore gradient, and derive indications for monitoring design. Under the Marine Strategy Framework Directive, eighteen stations at 0.5, 1.5 and 6 nautical miles from shore were sampled twice yearly with a Manta net from 2015 to 2025 (329 samples). Particles of 0.3–5 mm were isolated by sieving and identified by stereomicroscopy. Log-transformed concentrations were analyzed with linear mixed-effects models accounting for samples from the same outing, comparing an extended set of candidate models. Year was the only supported factor: model-based concentrations peaked in 2016–2017 and were about nine times lower in 2020–2021 (95% CI 5.4–14.7), whereas sites, distance bands and surveys did not differ and no site-specific trajectories emerged. The median concentration (0.044 n m⁻²) was comparable with other Italian coastal waters, and half of the unexplained variability arose between sampling outings. Contamination varies non-linearly between years rather than progressively. Monitoring precision would gain more from additional outings than from additional stations per outing.