Maximal tumor diameter (MD) is traditionally an important prognostic factor in breast cancer. It must be questioned, however, how well a one-dimensional parameter alone can represent the actual morphologic condition of a three-dimensional body. Along with the pathologically assessed MD and two perpendicular diameters (PDs) of a lesion, eccentricity (EF) and the three-dimensional parameters tumor volume (TV) and surface area (TSA) of 395 ductal invasive breast carcinomas of limited size (10-40 mm) were calculated. The dependent prognostic variable was axillary lymph node involvement (ALNI). MD, TV and TSA area were highly significant predictors of ALNI; these variables had similar levels of prediction accuracy (univariate analyses: MD: P = 0.0003, TV: P = 0.0009, TSA: P < 0.0001; multivariate analyses: MD: P = 0.0018, TV: P = 0.0109, TSA: P = 0.0009; pseudo R-squared values: MD: 0.42, TV: 0.39, TSA: 0.39). Despite certain variations in tumor shape, TV and TSA with similar MD, there is no evidence that three-dimensional pathologic measurements (TV/TSA) are more precise prognostic predictors of ALNI compared to the one-dimensional measurement alone.