This study aims to determine the effect of the use of loose parts media on improving the science ability of children aged 5-6 years. Research method quasi-experimental quantitative method, with non-equivalent design one group design by comparing pre-test and post-test values. Sampling using a saturated sample with a total of 15 students. Data was collected with an observation sheet in the form of a checklist. The collected data were analyzed with descriptive statistics and inferential statistics of paired t tets sample tests. The results of data analysis showed that the mean value of the postest was higher than the mean value of the pretest. The hypothesis test shows that the calculated t value is greater than the table t value with a significance value of 0.00, so it can be concluded that the use of lose part media can improve students' science skills.