Objectives: Volume measurement of subcutaneous xenograft tumors in nude mice models is an important metric to assess tumor growth or response to therapy. Manual calipers are widely used for this purpose. But the measurements with manual calipers may be inaccurate. Contrarily, three-dimensional (3D) ultrasonographic measurements give reliable and accurate tumor volume calculation. We aim to; evaluate the accuracy of common four formulas given in the literature to estimate xenograft tumor volumes based on manual caliper measurements and offer a new coefficient for a better estimation of the tumor volumes. Patients and Methods: Detailed manual diameter measurements of xenograft tumors were in 14 nude mice performed using Vernier caliper. Tumor volumes were calculated using the suggested formulas in the literature based on manual measurements. 3D ultrasound volume measurements were performed on same xenograft tumors using high resolution Vevo 2100 imaging system. To propose a new coefficient; means of ratio between manual and ultrasound volume measurement values were used. Also, data set was divided into two subgroups as tumor volume under 800 mm3 and over 800 mm3. New coefficients for each subgroup were defined. Results: Only with prolate ellipsoid formula there was no statistically significant difference between volume measurements with two methods (p=0,24). Our proposed formula (0,45 L*W*H) could estimate tumor volumes as good as prolate ellipsoid formula. Coefficient 0,35 and 0,81 in the same formula were found efficient in the selected subgroups. Conclusion: Using one common coefficient/formula for tumor volume estimation in any tumor size can be inaccurate. Appropriate coefficient should be chosen according to the dataset worked with.