Scatter dose may be calculated by summing the scatter contribution from individual volume elements. These contributions may be represented by differential scatter-air ratios (dSAR). Determination of dSAR from measured data is only approximately correct for second and higher orders of scatter and yields values often limited to one significant figure. Monte Carlo calculation, on the other hand, is time intensive, requires some knowledge of the beam's x-ray spectrum, and mastering the complexities of a program such as EGS4. Total scatter dose at a point may be determined by measuring depth dose or tissue-air ratios and partitioning the dose into its primary and scatter components. Scatter may be represented by scatter-air ratios, which can be characterized by the sum of first, second, and higher orders of scatter. The first scatter dose may be computed exactly by summing the first scatter contribution from individual elements, determined from the first principle. Separation of dSAR into primary attenuation and depth-independent terms allows the latter to be precomputed once for a given energy and stored in tabular form. Second scatter may be treated in a similar manner. The higher orders of scatter are computed by subtracting the sum of calculated first and second scatter doses from the total scatter dose. Elements close to and approximately 1 cm above the point of calculation contribute most heavily to the first scatter dose. Compared to the first scatter dose, the second scatter dose contribution is lower, particularly for elements close to the point of calculation.(ABSTRACT TRUNCATED AT 250 WORDS)