This study was aimed to compare the reproductive performance of Jawarandu goat which serviced naturally and artificially. In total 433 Jawarandu goat ewes were randomly allocated into two main groups namely Natural Mating (NM; n=233) and Artificial Insemination (AI; n=200). In this study, the reproductive performance was measured in the percentage of pregnancy, litter size, and percentage of kid born single, twin and triplet. Between groups, data were statistically compared using t-test at α = 5%. Natural mating group was performed in mating pen, and per Boer buck goat was colonized with 20-25 ewes for 45 days, followed with pregnancy check using ultrasonography (USG). In AI group, oestrous signs were observed in morning and evening, while insemination was done maximum 12 hours after the sign of oestrous using Boer frozen semen. Later, USG was applied to test the pregnancy at day 45 post insemination. The percentage of pregnancy in NM was significantly higher (p < 0.05) compared to AI group (73.39 vs 21.00%). However, there was no difference in litter size parameter (1.37±0.49 vs 1.35±0.56; p>0.05). Similarly, in the percentage of kid born single, twin and triplet, no differences were found in both groups (63.35 vs 64.86; 36.02 vs 35.14; and 0.62 vs 0%; respectively). All in all, natural service gives better reproductive performance compared to the artificial one. Its suggested due to factors related to the ability to determine the right time for service. Others factor might involve, such as the oestrus observation time and operator experience, especially in observing oestrus of ewe which didn’t show observable oestrus sign.