Twelve English- and 12 Urdu-speaking males, ranging in age from 18 to 30 years, were asked to assign pure-tone frequencies as a "best fit" to visually presented shapes. Six basic figures were employed (circle, ellipse, right triangle, isosceles triangle, along with two historical psycholinguistic forms-uloomu and takete) and varied on three dimensions of size, complexity, and density. The results of the study indicate that there was consistency in the assignment of pure-tone frequencies to the dimensions of the figures and that the pattern of assignment was generally similar for both English- and Urdu-speaking subjects alike. Specifically, for both language groups, round figures (circles and ellipses) were assigned significantly lower frequencies than other stimuli. Also, complex and dense figures (i.e., those designed to simulate "visual depth" and "visual texture," respectively) received significantly higher frequency settings than those stimuli not possessing these dimensions. Additional analyses revealed several factors that may have contributed to the observed pattern. For example, Urdu subjects of high English proficiency produced frequency assignments that closely resembled those of the English-speaking subjects, while subjects of low English proficiency produced a pattern that was somewhat different, particularly with regard to round, complex figures. Length of residency in the United States (6 months, 1 year, or 2 years), however, did not significantly influence frequency assignments for the Urdu subjects. These results are discussed in terms of "cross-cultural" versus "language-specific" phonetic symbolism.