If the TDS correction of Bragg reflexions is evaluated by integrating the TDS interference function over the range of measurement, the correction is usually, as Cochran has pointed out, too large. The magnitude of the 'overcorrection' is difficult to estimate since the proper evaluation of the correction involves the convolution of the interference function with the resolution function of the experimental set-up. Thus, in general, a sixfold integration is involved. Here calculations performed with a simple model which contains the essential features of the problem, and for which the integrations can be carried out properly, are reported. It has been found that the size of the 'overcorrection' depends only a little upon the absolute size of the range of measurement but strongly upon the size of the range of measurement relative to the magnitude of instrumental broadening (Bragg peak). With decreasing range of measurement relative to the Bragg peak the 'overcorrection' increases rapidly. For typical experimental situations of single-crystal measurements the 'overcorrection' seems to amount to about 5 to 20% of the TDS correction.