A low-cost adsorptive composite was prepared using natural zeolite, clinoptilolite, and shrimp shell wastes as raw materials. The shrimp waste was used to synthesize chitosan by the deacetylation of chitin which was extracted from the exoskeleton of crustacean shrimps. Response surface methodology (RSM) was employed to model and optimize the effect of process variables including initial dye concentration (25–100 mg/L), solution pH (3–11), adsorbent dose (1–3 g/L), and contact time (15–180 min) on the percentage of dye removal as the response. The optimum levels of independent variables for the removal of Reactive Red 120 (RR120) were time of 138.75 min, initial concentration of 43.79 mg/L, adsorbent dose of 2 g/L, and pH of 5. The optimum values for the removal of Reactive Red 196 (RR196) were time of 125.77 min, initial concentration of 44.72 mg/L, adsorbent dose of 2.5 g/L, and pH of 5. Under optimal conditions, the predicted removal percent of the model was 91.2 and 100% for RR120 and RR196, respectively. The results showed that the experimental data of RR120 were better described by Hill and Sips isotherms, and experimental data of RR196 followed Jovanovic isotherm model. The pseudo-first-order and pseudo-second-order models were employed to analyze the kinetic data. The kinetic data were well described by pseudo-second-order model with R Adj 2 values of higher than 0.98 and 0.99 for RR120 and RR196, respectively. The results showed that all of the independent variables exhibited a significant effect on dye removal.