Usability testing is not the numbers game that some people expect it to be. A relatively small number of test participants is usually sufƒcient to generate accurate and useful ƒndings. One widely recognized study suggests that just ƒve test sessions will generate many of the ƒndings that could result from a much larger study. Medical device regulators seem comfortable with summative usability tests involving 15-25 participants, presuming a reasonably homogeneous user population.