Secchi disk transparency in lakes is a function not only of the turbidity caused primarily by plankton, but also of the organic color level in the water. Multiple regression analysis of data from 55 Florida lakes yielded a close-fitting equation of the type (SD)−1 = a(Color) + b(Turbidity) + c. The statistical relationships between inverse transparency and chlorophyll a and between log transparency and log chlorophyll a exhibited greater scatter. Experiments in which concentrated solutions of humic substances were added to a large plastic "limno-bag" verified the linear relationship between (SD)−1 and color content but yielded a different slope than that obtained by regression analysis. The maximum possible transparency (assuming zero turbidity) was similar for both the experimental and regression relationships at color levels above 50 platinum (Pt) units, but increasingly divergent results were predicted by the two approaches at lower color levels. At a color of 100 Pt units, the maximum Secchi disk transparency is about 2.4–2.8 m. Because of the crudeness of transparency measurements, use of the above regression equation to compute transparencies from measured turbidities appears to be more reliable than use of the experimentally derived equation. Key words: Secchi disk, transparency, color, turbidity, trophic indicator, chlorophyll a, light attenuation