The widespread presence of plastic contamination in the environment presents a severe threat to human and animal health. This study introduces a toluene dispersion strategy for detecting microplastics of different sizes using surface-enhanced Raman spectroscopy (SERS). The evaporation-induced self-assembly (EISA) method was employed to prepare SERS substrates by incubating silver nanoparticles (AgNPs) of ∼40-60 nm with a microplastic solution containing polystyrene (250 μm, 2.1 mm), polypropylene (10-50 μm), and polyvinyl chloride (1-5 μm). SEM images and Raman spectroscopy confirmed the uniform decoration of AgNPs on filter paper substrates, with a relative standard deviation (RSD) of 8.22%. Thiophenol was used as a Raman reporter to monitor surface changes, showing a strong correlation (R2 = 0.986-0.995) between its SERS signal and microplastic concentration in aqueous and real samples. This is the first time a toluene dispersion strategy has been integrated with EISA to achieve highly sensitive microplastic detection, reaching a limit of 0.001 mg mL−1. The method was validated in real-world matrices, including lake water and salt samples, in the presence of interferents such as organic pollutants, inorganic ions, colloids, bio-organisms, and bisphenol A. This approach enables rapid detection of diverse microplastics in complex environmental samples.