BACKGROUND AND PURPOSE: Many patients with acute stroke are excluded from receiving thrombolysis agents within the necessary time limit (3 or 6 hours from stroke onset) because they or their family members are unable provide the time of stroke onset. Brain tissue sodium concentration ([Na(+)]) increases gradually and incessantly during the initial hours of experimental focal cerebral ischemia but only in severely damaged brain regions. We propose that this steady increase in [Na(+)] can be used to estimate the time after arterial occlusion in the rat middle cerebral artery occlusion model of ischemic stroke. METHODS: Sixteen anesthetized Sprague-Dawley rats underwent permanent middle cerebral artery occlusion combined with bilateral common artery occlusion. After 100 to 450 minutes, diffusion-weighted MRI was used to generate apparent diffusion coefficient (ADC) maps, cerebral blood flow (CBF) was determined with (14)C-iodoantipyrine (in a subset of 7 animals), and the brain was frozen. Autoradiographic CBF sections and punch samples for Na(+) analysis were obtained from the brain at the same level of the MR image. Severely at risk regions were identified with an ADC of <520 microm(2)/s and, in the subset, with both ADC of <520 microm(2)/s and CBF of <40 mL. 100 g(-1). min(-1). RESULTS: Both CBF and the ADC dropped quickly and remained stable in the initial hours after ischemic onset. Linear regression revealed strong linearity between [Na(+)] and time after onset, with a slope of 0.95 or 1.00 (mEq/kg DW)/min, with both ADC and ADC-plus-CBF criteria, respectively. The 95% CIs at 180 and 360 minutes were between 41 and 52 minutes. CONCLUSIONS: The time after ischemic onset can be estimated with this 2-step process. First, ADC and CBF are used to identify severely endangered regions. Second, the [Na(+)] in these regions is used to estimate time after onset. The favorable 95% CIs at the time limits for thrombolytic therapy and the availability of measurements of ADC, CBF, and [Na(+)] in humans through the use of MRI suggest that this time-estimation scheme could be used to assess the appropriateness of thrombolysis for patients who do not know when the stroke occurred.