In process automation, skill decay has not been investigated systematically. In two experimental studies, refresher interventions (RIs) were compared to support skill and knowledge retention of a start-up procedure on a simulated process control task. Based on theories of proceduralisation, the 'testing-effect' and mental practice, four different RIs were designed. In Study 1 (N = 68), two experimental groups (EGs) received either an RI called 'Practice' or an RI called 'Skill Demonstration'. Both RIs support skill retention, but the Skill Demonstration-RI performed with a higher mental workload. In Study 2 (N = 68), two EGs received an RI called 'Symbolic Rehearsal' or an RI called 'Procedural Knowledge Test', and aimed at supporting knowledge retention. Both EGs supported knowledge retention but showed moderate skill decay. Results imply that RIs affect skill and knowledge retention differently and should be applied in accordance with the task requirements and their dependence on accurate skill or knowledge maintenance.