Microplastics have been observed in indoor and outdoor air. This raises concern for human exposure, especially should they occur in small enough sizes, which if inhaled, reach the central airway and distal lung. As yet, methods for their detection have not spectroscopically verified the chemical composition of microplastics in this size-range. One proposed method is an automated spectroscopic technique, Raman spectral imaging; however, this generates large and complex data sets. This study aims to optimize Raman spectral imaging for the identification of microplastics (≥2 μm) in ambient particulate matter, using different chemometric techniques. We show that Raman spectral images analyzed using chemometric statistical approaches are appropriate for the identification of both virgin and environmental microplastics ≥2 μm in size. On the basis of the sensitivity, we recommend using the developed Pearson's correlation and agglomerative hierarchical cluster analysis for the identification of microplastics in spectral data sets. Finally, we show their applicability by identifying airborne microplastics >4.7 μm in an outdoor particulate matter sample obtained at an urban sampling site in London, United Kingdom. This semiquantitative method will enable the procurement of exposure concentrations of airborne microplastics guiding future toxicological assessments.