Seroma formation remains a significant clinical problem which increases morbidity and hospital costs in patients undergoing mastectomy operations. This study evaluated the effect of varying the concentrations of fibrinogen and thrombin in fibrin sealant on successfully preventing seroma formation in a rat model. After axillary dissection, control animals (Groups I and II) had 1 ml of either normal saline or thrombin (100 U/ml) applied to the axilla while treated animals (Groups III-VIII) had increasing concentrations of 0.5 ml of fibrinogen (25, 50, or 100 mg/ml) and 0.5 ml of thrombin (10 or 100 U/ml) applied. Seroma volumes (means +/- standard deviation) were measured on Postoperative Day 5. They were largest in Group I (3.1 +/- 1.9 ml, n = 13) and Group II (3.9 +/- 2.7 ml, n = 15) and then decreased from a high in Group III (2.5 +/- 2.4 ml, n = 15) using fibrinogen, 25 mg/ml, and thrombin, 10 mg/ml, to a low in Group VIII (0.8 +/- 1.0 ml, n = 15) using fibrinogen, 100 mg/ml, and thrombin, 100 U/ml. Analysis of variance revealed a statistically significant difference in the mean seroma fluid volumes between the groups (P = 0.0021), while Scheffe's comparison showed a specific significant difference (P = 0.028) between the thrombin control (Group II) and the highest concentration fibrin sealant (Group VIII). The difference in seroma volumes for all control animals (Groups I and II), 3.5 +/- 2.4 ml, and all treated animals (Groups III-VIII), 1.7 +/- 2.1 ml, was highly significant by unpaired t test. (P < 0.0001). Thus, fibrin sealant was useful in reducing seroma formation in this rat model with the highest concentration of fibrinogen and thrombin appearing most effective. These data may be useful in guiding future clinical trials in humans.