Previous epidemiological studies have indicated the effectiveness of gardening and other activities for dementia prevention. Our previous study using near-infrared spectroscopy (NIRS) suggested seeding and watering tasks could induce greater activation than other gardening tasks in the prefrontal area of older adults, but repetitive effects were not fully addressed. The objective of this NIRS study was to investigate the effects of repeated gardening tasks on activation of the frontal pole (FP). We measured oxygenated hemoglobin (Oxy-Hb) values in the FP while 24 healthy right-handed older adults (60 -73 years) performed a seeding task, a watering task, and a motor programming task (FAB 3) of the Frontal Assessment Battery (FAB). Each task was repeated five times. After the fourth trial, comparison of maximum Oxy-Hb values among the three tasks demonstrated that, in the medial FP, activation during the two gardening tasks was significantly or marginally significantly greater than activation during the FAB 3 task. The results may be attributable to differences in the number of sensory information sources and the number of stimulus-oriented (SO) attention switches and SO thoughts. No significant differences were observed in activation in the lateral FP. Comparison of maximum Oxy-Hb values among five trials within each task showed sustained medial FP activation in all tasks, while the watering task and FAB 3 task showed significantly decreased activation in a portion of the lateral FP. This was possibly due to differences in the number of times switching between SO attention and stimulus-independent (SI) attention, and differences in frequency of changes of visual dimension weighting. Also, in the gardening tasks, the participants needed to respond to changes in the external environment; it can be speculated that the need for collating external information corresponding to changes in the external environment induced repeated FP activation during the gardening tasks.