In this work, competitive adsorption of four food dyes (Sunset Yellow, Allura Red, tartrazine, and Brilliant Black) by heated pine wood is studied using multivariate calibration. Partial least squares PLS1 and principal component regression PCR are effectively applied for simultaneous determination of food dyes with high accuracy (96.2–103.6%) and low relative prediction error (2.3–6.9%). Using multivariate calibration, the dyes are detected down to 0.11, 0.15, 0.24, and 0.29 mg L⁻¹. The removal of dyes increased at acidic solution and H-bonding is the main controlling mechanism. The maximum removal capacities (according to Langmuir model) are 3.4, 2.5, 4.8, and 7.1 mg g⁻¹ for tartrazine, Brilliant Black, Allura Red, and Sunset Yellow, respectively, at pH 2.0 and 25°C. The competition factors (CFs) are estimated from the isotherms to assess the degree of competition between dyes toward the surface. The CFs are 0.64, 0.66, 0.77, and 0.77 for Brilliant Black, Allura Red, Sunset Yellow, and tartrazine, respectively. Accordingly, Brilliant Black is the most affected dye while Sunset Yellow and tartrazine are the least affected in multi-solute adsorption. This study demonstrates the useful application of multivariate calibration for studying competitive adsorption of colored pollutants with minimum experimental efforts expenses.