A method was reported in a previous paper (Entwistle and Terrill, J. Mater. Sci. 35 (2000) 1675) for the measurement of the micro-fibril angle in the S2 layers of softwood. The small-angle X-ray scattering pattern was recorded with the beam directed at 45degrees to the radial direction in the cell structure. A cruciform scattering pattern was produced and the micro-fibril angle was deduced from the angle between the arms of the cross. The analysis assumed that all the cell walls lay in either the radial or the tangential direction. Real cell structures do not conform to this ideal. The objective of the work now presented was to calculate the error in the deduced value of the micro-fibril angle arising from the assumption that the specimen cell walls all lie in either the radial or in the tangential direction. To this end, the length and orientation of over 1000 cell walls was measured on two specimens using an image analyser. From these data the azimuth angle at the peak of the scattered intensity was calculated as a function of the micro-fibril angle and the standard deviation of the spread of intensity. The true micro-fibril angle was determined by comparing these data with the measured azimuth angle at the peak intensity. A value for the micro-fibril angle was also calculated from the measured value of the azimuth angle at the peak intensity using the relation for a square-section cell structure M = Atan(root2* tanphi). The values differed from the true values by between 1degrees and 2degrees for micro-fibril angles between 25degrees and 30degrees. For many purposes this degree of error is tolerable and in those circumstances the above relation can be used to extract a micro-fibril angle value.