Antibiotic residues have been found in environmental samples, such as water, soil, and even food, and usually come from wastewater, presenting environmental and human health risks. This study aimed to improve the elimination of the antibiotics tetracycline (Tet) and chloramphenicol (Chlor) by modifying three factors: contact time (3–7 days), plant biomass (10–14 g), and antibiotic concentration (5–15 mg/L Tet and 10–20 mg/L Chlor). An approach that optimizes time and resources, response surface methodology (RSM), was applied with a Box–Behnken design (BBD) to two plant species (L. gibba and A. filiculoides), i.e., one experimental design was used for each species. Antibiotic residues in water and plant samples were analyzed by liquid chromatography. The optimal conditions for Tet removal were 6.04 d, 11.4 g, and 13.4 mg/L with Lemna and 6.3 d, 11.9 g, and 14.7 mg/L with Azolla; the optimal conditions for Chlor removal were 7.8 d, 13.6 g, and 10.2 mg/L with Lemna and 4.6 d, 12.3 g, and 8.7 mg/L with Azolla. The results showed that the removal efficiency of antibiotics increased depending on the species used, reaching a maximum of up to 100%. Tet was better removed than Chlor, reaching maximum removal values of 100% and 84% with Azolla and Lemna, respectively. Chlor removal reached 70% and 64% with Azolla and Lemna, respectively. The mean bioconcentration factors (BCFs) of Tet were 2.9% in Lemna and 4.9% in Azolla, and the BCFs for Chlor were 38.1% in Lemna and 37.8% in Azolla. Thus, in general, better results were obtained with Azolla. In summary, the results demonstrate that this design and the selected plants contribute to the removal of antibiotics, presenting a sustainable and recommended alternative for the treatment of wastewater contaminated with antibiotic residues.