Because public opinion polls have become important sources of information for legislators and politicians, public opinion pollsters should carefully consider the relationships between gender wording of polling questions, measurement errors, and validity and generalizability of survey findings. Findings from numerous research projects indicate that wording of questions is a major factor influencing polling results. It is hypothesized that gender wording of polling questions is a major factor contributing to errors in public opinion measurement. Review of sociolinguistic literature reveals a growing controversy over use of the male pronoun to indicate females as well as males. Analysis of gender images in 3,358 questions asked by nine polling organizations from 1963-73 indicates that 10% of questions were gender discriminatory. Areas in which women are significantly underrepresented in the questions include politics, status concerns, and economic affairs. Findings indicate that questions did not change during the 30 year period of analysis to reflect increasing labor force participation by women. One implication of the research is that gender biased public opinion poll questions serve to perpetuate sex stereotypes and cultural values. (Author/DB)