The dichotic discrimination test for children could make an important contribution to the diagnosis of central auditory processing disorders. With previous test procedures and their interpretation it was not possible to get the auditory discrimination ability in only one grade. To evaluate this ability, it was necessary to measure the degree of comprehension in each ear as well as the required noise level. The evaluation cannot be used for statistical results because there are too many parameters. Moreover, results apparently similar can come from varying abilities. We are therefore working on a new method to evaluate dichotic hearing more accurately. By using a constant noise level and only evaluating positively those hearing pairs which have been repeated correctly, we arrive at one single grade. The dichotic discrimination test was administered to 46 children with auditory perception disorders and the results evaluated according to both, the previous and new evaluation system. According to the new evaluation, a maximum of 20 correct responses is possible. The average number of correct responses was significantly lower than by the previous evaluation system. The standard deviation increased from 16% on the left side and 13% in the right to 24%, or five word-pairs. The difference is even clearer when the test is subdivided into four performance ranges, each of 25%. Of all the children examined, 85% fell in the upper performance range according to the previous evaluation. No child was in the lowest performance range. According to the new evaluation, only 56.5% of the children were in the upper performance range, while three children, that is 6.5%, were in the lowest performance range. This difference, P = 0.0002, is highly significant. Our results demonstrate clearly that on the basis of the new evaluation system the measurement of dichotic ability is possible, for the first time. We consider the new evaluation method to be an improvement in determining dichotic discrimination and in comparing it with other abilities. To check the validity, further independent testing should be carried out and results compared with those obtained from children without auditory perception disorders.