When individuals in samples from two populations are classified according to the presence or absence of a particular characteristic, the data are often presented in the form of a 2x2 contingency table. In some cases, for example when the characteristic is "alive or dead", there is little risk of error in classification, but when the characteristic is more difficult to assess, such as the presence or absence of a disease, errors of misclassification are likely to occur. The author discussed the effects of misclassification on tests of significance and on estimation. In general, misclassification will not affect the validity of a significance test but it will reduce the power of the test. The effect on estimation is more pronounced and can cause substantial bias. Suggestions are made for handling such data and a table is presented which shows the effect of misclassification on the efficiency of a significance test.