Citation-count is widely used in assessing the impact of papers and how to mine the valuable papers quickly and accurately is an urgent exploration in bibliometrics. This paper is aimed to summarize systematically the indicators and prediction methods for scientific pa鄄 pers citation impact. According to the traditional bibliometrics and emerging webmetrics, all the factors are classified into paper factors, author factors, journal factors, webmetrics and other factors. Then the methods on predicting citation-count are discussed, focusing on re鄄 gression analysis. Finally, advices about selection of indictors and data sources are presented.