BACKGROUND Swim-up techniques for sperm separation may have detrimental effects on sperm DNA. We wished to determine whether the normal swim-up method with centrifugation used in our laboratory, which involves a centrifugation step, was harmful to sperm compared with swim-up without centrifugation. METHODS Semen samples were obtained from patients undergoing IVF or andrology assessment. An aliquot was removed for fixation and subsequent DNA fragmentation determination. The remaining sample was divided into two equal parts, which were subjected to swim-up either with (normal swim-up) or without (direct-swim-up) centrifugation. Semen analysis was performed both before and after swim-up. DNA fragmentation, in spermatozoa previously fixed in 4% paraformaldehyde, was assessed by the terminal transferase-mediated DNA end-labelling procedure (TUNEL). The percentage of spermatozoa with DNA damage after each swim-up technique was compared with that in the original semen sample. RESULTS DNA damage was <5% in most samples. No significant change in DNA fragmentation was observed between the two swim-up procedures, although the 'normal' swim-up sample had significantly less DNA fragmentation than the pre-swim-up sample. CONCLUSIONS We conclude that our normal swim-up technique caused no more DNA damage to spermatozoa from normal semen samples than a direct swim-up technique that involved no centrifugation step.