Cytological evaluation of lymph nodes (LN) in canine cutaneous mast cell tumors (MCT) has a key role in MCT staging. However, cytological discrimination between metastatic and reactive LNs is debated and diagnostic criteria inconsistent. The aim of this study was to retrospectively quantify nodal mast cells (MCs) in non-oncological (NOD) and MCT-bearing dogs (MCTBD), using different sample preparation techniques, to evaluate the significance of the MCT number. Cytological specimens from NOD-LNs (10 fine-needle aspirates—FNAs) and MCTBD-LNs (10 FNAs, 10 scrapings, 10 touch imprints) were evaluated. MCTBD-LNs were grouped in: non-metastatic, possibly-metastatic, and metastatic based on current literature criteria. MCs were counted in 4, 8, and 20 high-power-fields, and over 500, 1000, and 2000 total cells. MCs were significantly more numerous in MCTBD-LNs than in NOD-LNs and in “metastatic” samples than in “non-metastatic”. There was no significant difference between “metastatic” and “possibly metastatic” samples. Sample preparation techniques did not influence these results. A negative correlation between MCs number and sample cellularity was observed. Results were confirmed regardless of the counting method applied. MCs counting per se cannot distinguish possibly metastatic and metastatic cytological samples. Sample preparation technique and the counting method applied seem to have no influence on cytological quantification of nodal MCs in MCTBDs.