The objective of the research was to evaluate the coagulation process with starchy maize starch to remove iron, turbidity and colour from the confined aquifer water. The coagulation efficiency was determined experimentally using the jar test equipment, carrying out tests at 6 levels of applied dose (1, 2, 3, 4, 5 and 6 mL) and at 5 levels of agitation speed (100, 150, 200, 250 and 300 rpm.) with defined times (1.5 minutes for fast mixing, 15 minutes for slow mixing and 10 minutes for sedimentation), after preparation of the coagulant at 3 concentrations (0.5%, 1% and 2%). The data obtained were statistically analysed for two contexts: to evaluate whether there are significant differences in removal between pre- and post-treatment values in the 3 parameters studied, using the paired Student's t-test; and to determine the most efficient treatment using the complete coagulation process (coagulant concentration, applied dose and mixing speed/time conditions), for which Duncan's test was used for a completely randomised design (CRD). As a result, it was obtained that when using a dose of 50 mg/L (ratio: weight/volume) at a concentration of 1% of solution, the appropriate agitation speed is 250 rpm for fast mixing and 63 rpm for slow mixing, as it has allowed an iron removal of 80% (0.55 mg/L), turbidity of 88.83% (4.94 NTU) and colour of 67.39% (50 Pt-Co).