We compared the effects of horticultural activities according to cognitive demand levels on psychophysiological responses in adults. Thirty-two adults in their 20s were included. Participants performed 10 horticultural activities (raking, planting transplants, fertilizing, tying plants to stakes, harvesting, sowing, conducting cuttage, planting potted plants, cutting and washing, arranging flowers) for 150 seconds at two levels of cognitive demand. Electroencephalographic (EEG) and electrocardiographic measurements were acquired during the activity. After each activity, the participants’ emotional states were evaluated using the semantic differential method (SDM). The EEG results, according to comparison by activity, showed that for nine activities, excluding raking, relative theta decreased when performing tasks at a level of high cognitive difficulty (HCD) compared with those with a low cognitive difficulty (LCD), and relative beta, relative gamma, fast alpha, and relative low beta increased, indicating activation of the prefrontal cortex. In the relative theta power spectra, the cuttage activity was found to be the lowest when performing tasks at a high level high cognitive difficulty, and the working memory function was activated the most compared with other activities. When sowing at a low level of cognitive demand, participants’ heart rate decreased and stabilized. When potted plants were harvested at a high level of cognitive difficulty, the ratio of low frequency to high frequency increased, and the sympathetic nervous system was activated. In addition, when planting transplants, and cutting and washing were performed at a high level of cognitive difficulty, and the standard deviation of the RR interval was high, indicating a high ability of the autonomic nervous system to resist stress. As a result of the SDM, the emotional state according to task difficulty was found to be more stable and relaxed than high cognitive difficulty, but a significant increase in comfort, relaxation, and naturalness was achieved when nine gardening tasks with low cognitive difficulty were performed, with the exception of sowing. The results of this study show that tasks with high cognitive difficulty activate working memory, whereas those with low cognitive difficulty stabilize and relax brain activation. Therefore, this study confirmed that an intervention in horticultural activities with an appropriate level of cognitive difficulty could have a significant effect on psychophysiological changes in adults.