Abstract: Abstract: The rapid method to evaluation of controlled release fertilizer had been established in China, which included the eletrical conductivety (EC) method and the refractive index(RI) method.The RI method was only applied for polymer coated urea. The substrates of the coated ureas were organic chemicals, however, during the the incubation, the urea release could be transformed into inorganic form such as NH4+, which would influence the test accurateness of the RI method. It was possible to incubate a working curve companied with tested samples would cancel the effect of urea form change. So 15 self-made coated ureas with different longevities were taken as samples, which were divided into three groups according to their nutrient release rates, and the samples were incubated in still pure water under the constant temperture of (25±0.5)℃ with the fertilizer/water ratio of 1/5, at the same time standard concentration urea solution was incubated under the same condition to get a working curve. And the urea incubation solutions were collected after 1, 7, 14, 21, 28, 42, 56, 70, 98, 126, 154, 210, 266 days during the incubation with renewal of pure water. After that the routine determination of H2SO4 degestion method and the RI method with a working curve at the same incubation interval companied with the tested samples were applied to test nutrient release rate of the coated urea, further more the Richards equation was applied to make a nonliner analysis for each nutrient release curve and calculate the longevities of coated ureas. Finally, the results tested by different methods were used to analyse their similarities and differences and offer a more accurate and rapid determination for polymer coated urea.The results were as follows. Firstly, the EC values of a series of standard concentration urea solution increased with the incubaiton period prolonged, but urea mass concentration and their RI values always had a significant linear correlation relationship (r2≥0.998) after 0, 1, 6, 7, 14, 21, 28 and 56 days during the incubation. However, the errors tested of the fresh working curves were from 0.71 to 4.85 g/L, which would be decreased to the range of 0.34 to 1.13 g/L by using working curves at a same incubation interval companied with samples. Secondly, compared to the routine determination of H2SO4 degestion, the data of nutrient release rate of coated ureas were lifted more than 10% by using the RI method with the fresh working curve, but the new RI method based on the working curve at a same incubation inerval would cancel the difference of results, which had a conversion coefficient of 0.990 with the traditional one's. Thirdly, in order to further analysis the accurateness of the new RI method, the Richards equation was applied to model nutrient release curve and calculate the longevity of each sample whose time point to nutrient release amount getting 75%. Results showed that the longevities tested of 15 coated ureas were 21.9-245.9 days and 22.0-210.1days respectively by using the rountine and new method, the minimum and maximum deviation of the new method was 0.8% and 25.1% respectively. The probable reason was that, when the coated ureas with a longer longevity getting more than 180 days were taken as samples, because of nutrient being released too slowly during the later stage, it was easy to bring the greater deviation tested, and their errors were from 27.6 to 40.4 days, but it was acceptable according to routine to label the longevity of controlled release fertilizer and permitted to move up or down about two months. In summary, it was proved that the RI method based on a working curve at the same incubation companied with tested samples was a rapid determination for coated urea, and the longevity would be forecasted accurately by using the Richards equation. Furthermore, the result of this study could offer a reference of a rapid and accurate method to test the nutrient release rate and longevity of polymer coated urea.