We compared intracranial pressure (ICP) and cerebral blood flow (CBF) in dogs after inflating a subdural intracranial balloon to increase ICP to 20 mm Hg, inducing hemorrhagic shock (mean arterial pressure [MAP] of 55 mm Hg), and infusing a single bolus of fluid consisting of either 54 mL/kg of 0.8% saline (SAL), 6 mL/kg of 7.2% hypertonic saline (HS), 20% hydroxyethyl starch (HES) in 0.8% SAL, or a combination fluid (HS/HES) containing 20% HES in 7.2% saline. Twenty-six dogs were ventilated with 0.5% halothane in N2O and O2 (60:40 ratio). As ICP was maintained at 20 mm Hg, rapid hemorrhage reduced MAP to 55 mm Hg (time interval of zero [TO]) which was maintained at that level for 30 minutes (until T30). Subsequently, over a 5-minute interval (T3O-T35), one of the four randomly assigned resuscitation fluids was infused. Data were collected at baseline; after subdural balloon inflation; at TO, T30, T35, and 30-minute intervals thereafter for 2 hours (T65, T95, T125, and T155). CBF and ICP were compared using repeat-measure ANOVA. Cerebral blood flow was greater at T35 in the HS and HS/HES groups than in the HES group (P =.025). In the SAL group, ICP increased significantly from TO ot T35, remaining unchanged thereafter. At T35, ICP in the HS group was significantly lower than in the SAL group (P <.05) but subsequently increased. ICP in the HS/HES group exceeded that in all other groups at T95 and T125 (P <.05). After a severe reduction in cerebral perfusion pressure (CPP), HS solutions (both HS and HS/HES) were associated with a delayed rise in ICP and did not improve global forebrain CBF in comparison with conventional saline solutions.