The accuracy of computer-assisted sperm morphometry analysis (CASMA) depends on the careful preparation, fixation and staining of spermatozoa. The efficiency of CASMA may be enhanced by developing optimized protocols. The aim of the present study was to evaluate the influence of sperm washing and the use of three staining techniques [rapid Panoptic, Hemacolor and Harris's Haematoxylin (HH)] on image-processing accuracy and boar sperm head morphometry. Sperm washing had a significant effect on samples stained with rapid Panoptic, increasing the percentage of correctly binarized sperm heads and the contrast between cells and background. However, rapid Panoptic yielded the lowest percentage of properly digitized sperm heads. HH provided the highest cell/background contrast, and also greater sperm head staining intensity, but discrimination of sperm midpieces was considered insufficient. Hemacolor occupied an intermediate position, providing acceptable colour intensity and satisfactory cell/background contrast. Use of different staining procedures prompted dimensional differences in sperm head morphometry. Significant differences between animals were observed for all morphometric parameters. Low within-animal variation coefficients reflected a homogeneous sperm head population. Between-animal variation coefficients were relatively high for Hemacolor and HH, and significantly high for the rapid Panoptic stain. Using Panoptic and HH, stable morphometric measurements required at least 100 properly digitized sperm heads rather than 200, while Hemacolor required only 50 spermatozoa. These results indicate that both washing of semen and staining procedures significantly affect the accuracy of image processing and sperm head dimensions. Hemacolor and HH proved to be the best staining techniques for evaluating sperm head dimensions in boar.