Biomonitoring using vegetation and epiphytic lichens has gained prominence as a low-cost and integrative approach to assess air quality impacts on ecosystems. This study aimed to evaluate the bioaccumulation and physiological responses of Mangifera indica leaves and epiphytic lichens across nine monitoring stations in southeastern Brazil, and to develop a predictive environmental risk index based on pollutant concentrations. Heavy metals (Cr, Cu, Zn, Mn) were quantified in soil, leaf, and lichen samples, and Air Pollution Tolerance Index (APTI) was calculated using foliar biochemical parameters. Data from São Paulo State Environmental Agency (CETESB) air quality stations, including PM10, NO, NO2, and NOX, were integrated using Principal Component Analysis (PCA), Canonical Correlation Analysis (CCA), and linear regression modeling. PCA revealed that Cr, Cu, and Zn in foliar and lichen tissues, along with APTI and fluoride, accounted for most of the biological variation. CCA identified nitrogen oxides (NO, NOx) as the primary environmental drivers of biological response. A robust regression model (based on CCA results) was developed to predict a composite biological risk score using air pollution data alone. This model was u91,sed to construct an Environmental Risk Index that accurately classified sites into high, moderate, or low-risk categories. The approach demonstrated here offers a transferable, biologically grounded framework for assessing environmental risk, particularly in urban-industrial contexts, and supports the integration of biomonitoring into environmental and public health policies. Keywords: Ecotoxicology; Environmental health; Bioindicators; Risk toxicity; Heavy metal; Bioaccumulation.