This paper describes the effect of scanning patterns and number of ima
ges in each scan on the accuracy of automated image analysis technique
s used for maceral analysis. The results are compared with the accurac
y of manual point counting. Current theoretical standards are discusse
d together with actual results obtained from within the laboratory. Wa
ys of increasing the accuracy of automated maceral analysis using diff
erent methods of sampling with automatic staging is demonstrated, toge
ther with the number of images which need to be scanned.