In a laboratory simulation of the activated sludge process ten heavy metals were added continuously to the system which was allowed to equilibrate at six sludge ages between 3 and 18d. Cobalt, manganese and molybdenum removals were poor and were unaffected by changes in the sludge age. The highest removal efficiencies for the other metals occurred at the 15d sludge age. Chromium (trivalent) and cadmium had the highest removal efficiencies, typically greater than 50%. The behaviour of the majority of the metals which were removed to a significant extent was related to one of the parameters influenced by sludge age, i.e. mixed liquor suspended solids, effluent suspended solids or effluent chemical oxygen demand. The metals which were poorly removed showed little affinity for the activated sludge, while most metals exhibited maximum specific uptake by the mixed liquor at a sludge age of 9–12d. However, the affinity of silver for the mixed liquor continued to increase as the sludge age increased to 18d. An affinity series, based on an arbitrary measure of the specific accumulation of metals by the mixed liquor, indicated that chromium, cadmium and silver were most readily adsorbed by the activated sludge.