The Ca-Fe-La composite was synthesized by the one-pot solvothermal method and applied to the advanced removal of phosphate from aqueous solutions in a fixed-bed column. The effects of some important process parameters on the phosphate adsorption were examined. It was found that the maximum equilibrium loading was 121.1mgg−1 at an influent phosphorus concentration of 30mgL−1, flow rate of 2.5mLmin−1, adsorbent mass of 0.5g and pH of 5.14. The Thomas and fractal-like Thomas models were used to analyze the measured breakthrough curves. Their fitting quality was diagnosed by the adjusted coefficient of determination (Adj. R2), the root of mean squared error (RMSE) and the residual plot. According to the Thomas and fractal-like Thomas models, the breakthrough time and the saturation time were solved by the fzero command of MATLAB software. Akaike information criterion (AIC) and Bayesian information criterion (BIC) were adopted to further determine the optimal model. In all cases, the phosphate adsorption on the Ca-Fe-La composite followed the fractal-like Thomas model. The phosphate loaded Ca-Fe-La composite could be used as a slow-release fertilizer and promote tomato plant growth. This work aimed to provide reliable assessment methods for the modeling of the breakthrough curves and allow for waste management and resource utilization of the phosphate loaded Ca-Fe-La composite.