Motor imagery (MI) and action observation (AO) have been found to enhance motor performance, but recent research found that a combination of action observation and motor imagery (AOMI) together is even better. Despite this initial finding, the most effective way to combine them is unknown. The present study examined the effects of synchronized (i e., concurrently doing AO and MI), asynchronised (i.e., first doing AO then MI), and progressive (first asynchronised approach, then doing synchronized approach) AOMI on golf putting performance and learning. We recruited 45 university students (Mage = 20.18 + 1.32 years; males = 23, females = 22) and randomly assigned them into the following four groups: synchronized group (S-AOMI), asynchronised group (A-AOMI), progressive group (A-S-AOMI), and a control group with a pre-post research design. Participants engaged in a 6-week (three times/per-week) intervention, plus two retention tests. A two-way (group × time) mixed ANOVA statistical analysis found that the three experimental groups performed better than the control group after intervention. However, we found progressive and asynchronised had better golf putting scores than synchronized group and the control group on the retention tests. Our results advance knowledge in AOMI research, but it needs more research to reveal the best way of combining AOMI in the future. Theoretical implications, limitations, applications, and future suggestions are also discussed.