Computer-aided sperm analysis (CASA) is now routinely used in IVF clinics, animal breeding centres and research laboratories. Although CASA provides a more objective way to evaluate sperm parameters, a significant number of factors can affect these measurements. This paper classifies these factors into four categories: (1) sample and slide (e.g. preincubation time, type of specimen and type of chamber slide); (2) microscope (e.g. light source and microscope stage); (3) hardware and software, including the settings of each system; and (4) user-related factors. We review the effects of the different factors in each category on the measurements made and emphasise the need to take measures to standardise evaluations. The take-home message of the present article is that there are several commercial and useful CASA systems, and all are appropriate for routine analysis. Non-commercial systems may also be good choices when the user needs to adapt the device to specific experimental conditions. In both cases (commercial and non-commercial), it is important that standard protocols are put in place for evaluation, as well as methods to validate the system.