There is no agreement as to which variable best mirrors disease activity in rheumatoid arthritis (RA) and no studies have been performed on the validity of disease activity variables. In this study the validity of 10 commonly used single variables and three composite indices was tested. All patients participated in a large follow up study in two clinics. The patients (n = 233) had classical or definite RA and a disease duration of less than one year at entry. The mean follow up time was 30 months; the follow up frequency was once every four weeks; 6011 records were used in the analysis. The validation criteria included correlations with the other variables (correlational validity), with the physical disability (criterion validity I), and with the radiographically determined damage of hands and feet (construct validity). The judgment of a group of rheumatologists in clinical practice was also used as a model of criterion validity (II). In this comparison the disease activity score and Mallya index showed the best validity. The best single variable was the number of swollen joints. The validity of most single variables was poor and these variables were not suitable as single endpoint measures in clinical trials.