This study aimed to non-destructively identify quality parameters in Javadi peaches. In this context, a capacitive array (CCD) spectroradiometer ranging from 350 to 1150 nm was employed for the nondestructive prediction of pH, titrable acid (TA), soluble solid-state (SSC), total phenol (TP), extract anthocyanin (Prance and Nesbitt). Utilizing multivariate partial least squares (PLS) regression models, both reference and degradation measurements were considered. Various spectral data processing techniques such as Savitsky-Goley smoothing (SG), first derivative (D1), incremental diffusivity correction (MSC), standard normal variate (SNV), and Baseline correction were applied individually and in combination (SG + MSC + D1) and (SG + Baseline) to forecast peach quality attributes. Model performance was assessed using the root mean square error of prediction (RMSEP), correlation coefficient (rp), and standard deviation ratio (SDR). Optimal models showed high accuracy for pH (raw spectra: RMSEP = 0.15, rp = 0.94, SDR = 3.0), moderate for TA (first derivative: RMSEP = 0.07, rp = 0.86, SDR = 2.29) and SSC (MSC: RMSEP = 1.17, rp = 0.85, SDR = 1.88), but lower for TP and EA (SDR < 1.5). This non-destructive approach offers rapid, cost-effective quality assessment compared to destructive methods.