Objective- To compare transabdominal (TA) and transrectal (TR) ultrasonography of the prostate gland. Design- Experimental study. Animals- 10 intact male dogs that required euthanasia for reasons other than prostatic disease. Procedures- Prostatic length (L), width (W) and height (H) were measured in T A (L1, W1 and H1) and TR (L2 and H2) ultrasonography. Prostatic volume was calculated using the formula for the volume of an ellipsoid (VE1 and VE2) or for a box (VB1 and VB2) in TA and TR procedures, respectively. The dogs were euthanized, the prostate gland removed, and actual prostate dimensions and volume were measured. Linear regression was used to compare prostatic dimensions and volume measured by ultrasonography to actual ones. Results- In TR ultrasonography there was no damage to rectum, the time for evaluation of the prostate gland was faster and the resolution of prostatic images was better than other procedure. There was highly significant correlation between prostatic parameters and their actual dimension. VB1, VE1 and VB2 were greater and VE2 were lesser than actual prostatic volume. In TR and TA methods the dimensions had significant correlation with real prostatic sizes, while had not significant statistical differences with real prostatic sizes. Conclusions and Clinical Relevance - TR ultrasonography found to be a simple, quick and noninvasive method and TA ultrasonography also yields useful information about the prostate. Calculated prostatic volume doesn't estimated prostatic volume correctly and need to use equations produced by linear regression: V = 0.427 VB1 – 0.88. V = 0.796 VE1 – 0.40. V = 0.680 VB2 – 1.95. V = 0.980 VE2 + 1.94.