In an attempt to assess the extent to which contamination by exogenous squamous cells contributes to false-positive results in the cytologic diagnosis of amniotic fluid embolism, blood sampled from the right heart or from the pulmonary artery of 42 nonpregnant cardiac patients was processed and examined for squames. Despite a variety of precautions, it proved impossible to eliminate contamination in any of these specimens. In addition, two of three blank slides that were stained with maximal precaution also demonstrated squames. The authors conclude that cytologic techniques are not helpful in the diagnosis of amniotic fluid embolism and should no longer be utilized.