Many educators and researchers who believe that girls have less mathematical ability than boys associate this deficiency with girls' inferior spatial ability. This generally accepted belief that spatial skills are a major prerequisite for math achievement was tested through a study of the visual-spatial skills of a sample of 32 professional mathematicians. Subjects responded to the Witkin Embedded Figure Test and participated in brief individual interviews regarding childhood environment and recreational activities. There was little indication from the test that a strong or clear causal relationship exists between the subjects' ability to solve problems presented in the EFT and their achievement in mathematics. Moreover, interviews indicated no consensus that visual-spatial skills were useful to mathematics. Researchers suggest that environment rather than innate ability may be a key factor in math performance. Thus, the source and utility of visual-spatial skills might better be determined by observation and examination of education and experience in school and elsewhere rather than by a search for innate, fixed, and immutable cultural and biological factors. Teachers' beliefs that girls have less math ability than boys or that girls are unsuited for math-related careers can seriously affect the way they teach and what their students learn. (LP)