This study aimed to determine the best concentration of sago liquid sugar to obtain the best quality of cookies. This study used a completely randomized design (RAL) with five treatments and three replications, namely: P0 (100% granulated cane sugar and 0% liquid sago starch sugar), P1 (75% granulated cane sugar and 25% liquid sago starch sugar), P2 (50% granulated cane sugar and 50% liquid sago starch sugar), P3 (25% granulated cane sugar and 75% liquid sago starch sugar), and P4 (0% granulated cane sugar and 100% liquid sago starch sugar). Data were statistically analyzed using ANOVA and DNMRT at a 5% level. The results showed that the ratio of granulated cane sugar and liquid sago starch sugar significantly affected chemical analysis and sensory tests. The result showed that the best treatment was P1 which had a water content of 6.38%, ash content of 0.86%, protein content of 0.96%, and reducing sugar content of 18.44% with the sensory description of the cookies were yellow, no sago sugar aromatic, sweet, and crunchy.