Indices for M-mode measurements in dogs usually have been based on the assumption that a linear relationship exists between these measurements and body weight (BW) or body surface area (BSA). The relationships between the geometry of 3-dimensional objects do not support this assumption. The purposes of this study were to retrospectively examine M-mode data from a large number of dogs of varying sizes and breeds that were examined by a large number of ultrasonographers, to use the allometric equation to determine the appropriate BW exponent required to predict these cardiac dimensions, and to determine normal mean values and prediction intervals for common M-mode variables. Linear regression analyses of data from 494 dogs (2.2-95 kg) revealed a good correlation between M-mode measurements and BW after logarithmic transformation of the data (r2 = .55-.88). Most variables were most closely related to an index of body length, BW(1/3), although the exponent that best predicted diastolic and systolic left ventricular wall thicknesses was closer to 0.25. No variable indexed well to BW or BSA. With these data, appropriate mean values and prediction intervals were calculated for normal dogs, allowing veterinarians to correctly and appropriately index M-mode values. The equations developed from this study appear to be applicable to adult dogs of most breeds.