Studies have used cell size as a proxy to estimate polyploidy levels, the number of chromosome sets in somatic cells, in modern and fossil plant species. This paper critically evaluates these methods by reviewing cell size- and polyploidy-related literature, and provides new cell size data from herbarium material and fossil remains of willow genus Salix. The 40 extant taxa used in the study include most of the polyploidy levels encountered in Salix (2n = 38, 76, 114, 152, 190). Diploid and tetraploid species were morphologically similar to the fossil specimens. Specimens from alpine and arctic regions, forms rarely found in the fossil record, were included to extend the range of polyploidy levels. Measurements taken for this study were on the petiole epidermal cells of fossil and herbarium specimens and the stomatal guard cell complexes of herbarium material. A literature review reveals cell size may not relate only to DNA content, but to a plant’s age, nutritional state and the seasonal timing of organ development. Cell size measurements show that cultivated plants grown at elevations more than 800 m below their original place of growth have a significant increase in cell size. Leaf length-to-width ratio, infrageneric classification, and adaptation to dry or humid environments also correlate with cell size. Cell size proxies for estimating polyploidy levels in fossil willows provide only accurate results if morphologically similar modern plant material from natural habitats is used as a reference for comparison. Leaves should be similar in overall shape, base and apex shape, blade length and width, length-to-width ratio, petiole length, petiole-to-blade length ratio, venation pattern, and margin dentition.