Surgical preparation of the cremaster muscle for microvascular studies typically requires disruption of collateral vessels within the muscle and between the cremaster and the structures of the epididymis/ductus deferens. To study the effect of interrupting these vascular connections on cremaster hemodynamics, two modified preparations were examined in addition to the conventional open cremaster muscle preparation. One of these preparations enabled the measurement of feed vessel (1A) pressure and diameter with all cremaster vascular connections intact. The second preparation involved interruption of intramuscle collaterals to open the cremaster sac but with intact collateral pathways between the cremaster and deferential vessels. Intravascular pressures in the main cremasteric arteriole (1A) were similar in all three preparations with pressure in the 1A, expressed as a percentage of femoral artery pressure, varying between 53% in the intact preparation and 48% in the standard open preparation. These data support the existence of substantial upstream vascular resistance regardless of the extent of surgery. Selective occlusion of the branches of the deferential vessels significantly increased red cell velocity in the cremasteric 1A and major draining venule (1V) so that calculated blood flow increased by approximately 40% (P less than 0.01) in the 1A and 95% (P less than 0.01) in the 1V. Also, intravascular pressure fell significantly (P less than 0.01) in the 1A and increased in the 1V. Despite these compensatory changes total blood flow to the muscle was reduced by approximately 40% in the standard open preparation, compared to the preparation with the deferential feed pathway intact. Further studies where the 1A flow was transiently occluded indicated that the deferential pathway was capable of providing significant collateral blood flow to the muscle. Collectively these studies demonstrate that the surgical modifications of the cremaster vascular supply required for in vivo microscopy significantly alter normal hemodynamics within the vascular bed. The surgery does not, however, entirely explain the large pressure drop that exists upstream of the cremasteric 1A.