When analyzing sunflower (Helianthus annuus L.) remains, which are often carbonized, archaeobotanists commonly differentiate between wild and domesticated achenes and seeds based on the measured length (L) and width (W) or the calculated index L*W. Carbonization reduces the dimensions. To compensate for these reductions, archaeobotanists use a single correction factor proposed by Richard Yarnell (1978) for all cases. The use of a single correction factor can bias the reconstructed dimensions as carbonization is a highly variable process. The current study determines the relationship between carbonization and the dimensions of length and width. Measurements established that a decrease of 2.5–22.5% in achene length and 10–29% in achene width can occur, depending on temperature, heating rate, and variety. For seeds, temperature is of most importance, and shrinkage ranges from 0–27% for the length and from 0–20% for the width. These ranges make the use of a single correction factor problematic. A method is developed in which reflectance (an optical property applied in coal technology to determine coal rank) is used to measure the carbonization temperature, and in turn the shrinkage can be calculated. Subsequently, correction factors are calculated to reconstruct the original length and width. When applied to an assemblage of carbonized sunflower achenes, the newly developed method shows that the Yarnell single correction factor may bias the dimensions towards classifications of “wild” or “ruderal” forms of sunflower.