Green tea (Camellia sinensis L.), one of the most popular beverages worldwide, contains caffeine, a natural stimulant. However, some green tea extracts have been known to possess both hypnotic and arousal effects. This study aimed to identify the components influencing these dual effects using a green tea ethanol extract (GE). Response surface methodology revealed that only some extraction conditions significantly induced arousal effects in ICR mice during the pentobarbital-induced sleep test. Among these, extraction with 95% ethanol for 195 minutes achieved the maximum arousal effect, corresponding to a caffeine content of 58.9 mg g-1, comparable to the effects observed with the reference, 25 mg kg-1 of caffeine. In addition, administration of this GE sample significantly increased wakefulness for 3 hours following treatment in C57BL/6N mice, as confirmed through sleep architecture analysis. A correlation analysis of the total phenolic content (TPC) to caffeine ratio in GE found that the intensity of the arousal-inducing effects varied with TPC (R2 = 0.9428). It was also confirmed that the ratio of EGCG to caffeine, major components of GE, was more closely associated with sleep duration (R2 = 0.9034). L-Theanine, known for its sleep-promoting effects, did not independently affect the arousal effects of GE. However, when combined with EGCG, their total content showed a slightly stronger correlation with sleep duration in relation to the caffeine ratio, compared with that of EGCG/caffeine ratio (R2 = 0.9464). Therefore, the balance between TPC and caffeine appears to modulate the stimulant properties of GE, highlighting its potential as both a stimulant and a mild hypnotic agent. Collectively, these findings provide insights into optimizing GE for tailored functional foods based on its polyphenol/caffeine ratio.