An attempt was made to evaluate the standard error of equating (at the mean of the scores) in an ongoing testing program. The interest in estimating the empirical standard error of equating is occasioned by some discomfort with the error normally reported for test scores. Data used for this evaluation came from the Admissions Testing Program of the College Board. The method used depends on the fact that about half the examinees take the Scholastic Aptitude Test (SAT) twice or more. The calculation of the standard error of equating SAT verbal and mathematical scores in this study makes use of the variance of the mean gains over the course of 12 to 17 years for which comparable data are available, separately by pattern of repetition. It is reasoned that because each of the means used to calculate a mean gain is based on data for a different form of the test, the variance of a number of these means would be attributable to some extent to the variation associated with equating error, and that the variance of the errors of the mean gains would equal the variance of the errors on the first occasion of testing plus the variance of the errors on the second occasion, or twice the variance of errors on either occasion. In the example, standard errors of equating were quite small. One table presents the standard errors. (SLD)