Background: Although the comet assay has potential value for measuring DNA damage in large epidemiological human sperm studies, it is impractical to perform the assay daily on fresh semen samples. Therefore, before its use in epidemiological studies, the reliability of the comet assay in measuring DNA damage in cryopreserved sperm should be compared with that in fresh human sperm. Methods: Semen samples from 16 men were cryopreserved in liquid nitrogen (LN) using four methods: flash freezing with and without cryopreservative, and programmable freezing with and without cryopreservative. Neutral microgel electrophoresis was performed and comets were stained with YOYO-1. Comet length was measured using an eyepiece micrometer at × 400 magnification. Results: The highest correlation was between comet assay results obtained from fresh human semen compared with semen flash frozen without cryopreservative (R = 0.88). However, the method of cryopreservation, as compared with other sources of variability, accounted for only 6% of the variability. Inter-individual variability accounted for 20%, and individual sperm-to-sperm variability within an ejaculate accounted for 65%. Conclusions: Flash-freezing in LN without cryopreservative most closely reproduced the results obtained using fresh human semen samples, and thereby represents the most appropriate cryopreservation method for human semen in epidemiological studies utilizing the neutral comet assay.