Methods for predicting the likelihood of development of rheumatoid arthritis for individuals that present with recent-onset undifferentiated arthritis. The methods are based on the determination of a set of clinical parameter values and determining a predicted risk for developing rheumatoid arthritis by correlating the parameter values with predefined risk values associated with ranges of parameter values. Parameters values that are decisive for the risk for developing rheumatoid arthritis may include serum levels of C-reactive protein, Rheumatoid factors, anti-CCP antibodies, as well as age, gender, localization of the joint complaints, length of morning stiffness, and number of tender and/or swollen joints. The method may be performed by a computer. The invention further relates to a computer, a sample analyser and a computer program product for performing the method and a data carrier with the computer program product.