The purpose of this study was to develop the cost-effective and efficiency seed longevity prediction method of rice (Oryza sativa L.) germplasm for viability monitoring. To find an optimum predicting method for rice seed longevity at genebank, an accelerated ageing (AA) test, a controlled deterioration (CD) test and a dry-heat treatment (DHT) were conducted to the four groups of rice germplasm based on ecotype, such as Indica, Japonica, Javanica and Tongil type. Among the three artificial aging treatments, the dry-heat treatment of 36 hours at 90°C is suggested as a routine predictive test method of rice germplasm longevity at a genebank. The distribution of germination rate on 3,066 accessions which conserved 26.5 years at 4°C showed similar trend with the result of distribution by dry-heat treatment at 90°C on 36 hours using 106 accessions of rice selected samples which composed four ecotype groups. The results show that the dry-heat treatment affect not only predicting the rice seed longevity but also determining effective interval for monitoring germination of rice germplasm in genebanks.