We have shown that astronomical records in Samguksagi, the book of history of Sinra, Paekche, Koguryŏ, are the results of independent observation of these three countries. We used two methods to prove this fact. The first is to show that astronomical records of the book not appearing in the history books of neighboring countries do correspond to real celestial phenomena. Eight independent records of "太白晝見" (the Venus appearing during the daytime) and seventeen independent records of close encounters of the moon with planets are compared with the results of celestial-mechanical computation. We found that many of them coincided with the real phenomena. Special attention has been paid to early "太白晝見" records of Paekche, ones in A.D. 224 and 394. The second way of proving the independence of the records is to show that the solar eclipses recorded by the three countries are of the nature significantly different from that of eclipses recorded by Chinese as a whole. It is found that Sinra's eclipses before A.D. 201 can be best observed in the area of the Yangtse river of the Chinese continent while those after A.D. 787 mainly passes the southern Korean Penninsula. The location where Paekche's eclipses can be optimally observed is found to be the region around Bohai gulf near Beijing. Koguryŏ's eclipses are located in the region with latitude much higher than those of Sinra's and Paekche's eclipses. We show from a set of Monte Carlo simulations that the probability that the early eclipse records in Samguksagi are mere random copies from Chinese history books is less than 0.026 percent. We have also shown that the region of East Asia having meteorological characteristics compatible with the flood records of Samguksagi is consistent with that found from eclipse records. We conclude that the astronomical records of Samguksagi are based on real observation performed by ancient Koreans. The chronology of the history book has also been confirmed by our calculation. The remaining problem, for historians, is that the locations consistent with early Sinra and Paekche's records of natural phenomena are not in Korean Peninsula.