Closed-loop experimental results for conversion control in the presence of reactive impurities in continuous solution methyl methacrylate polymerization reactors are presented. The desirable conversion level is maintained using initiator flow rate as the manipulated variable. During operation, known or unknown types and amounts of reactive impurities are injected into the reactor as disturbances. Conventional proportional-integral-derivative (PID) control algorithms (e.g. PID, Smith predictor and Dahlin’s control) and stochastic control strategies [e.g. minimum variance control (MVC), constrained MVC and one-step optimal control] are evaluated first at the simulation level to identify promising control runs. Then, closed-loop experiments follow in a series of two- to three-day continuous runs to verify the simulation results. This study triggers many interesting ideas for practical polymer reactor control.