We hypothesized that the evolution of cerebral edema accompanying cerebral hypoxia-ischemia is dependent on age and that such differences would be detectable using magnetic resonance imaging methods. Thus we examined in immature and juvenile rats the relationship between hypoxic-ischemic changes in T 1 and T 2 and the alterations in brain water content, as assessed by differences in tissue wet-dry weights. One-and 4-wk-old rats were anesthetized and subjected to unilateral carotid artery occlusion and subsequent exposure to hypoxia (8% oxygen). T 1 and T 2 maps were acquired at 9.4 T, and then brain water content was measured in sham controls or in hypoxic-ischemic animals before, during, and 1 or 24 h after hypoxia-ischemia. In sham controls, T 1 , T 2 , and proton density decreased with increasing age, corresponding to an ontogenic decrease in water content. In 1-wk-old rats, increases in T 1 and T 2 were observed during and at 1 and 24 h after hypoxia-ischemia, corresponding to elevations in water content. In 4-wk-old rats, T 1 and water content increased during and at 1 and 24 h after hypoxia-ischemia whereas T 2 was not increased until 24 h after hypoxia-ischemia. Regression analysis showed that T 1 correlated better with total water content than T 2 . In both immature and older brain, an increase in total brain water develops acutely and persists after an episode of cerebral hypoxia-ischemia, and T 1 imaging detects this change better than T 2 . Hypoxic-ischemic changes in T 2 are age dependent, reflecting other physicochemical changes of water in the tissue than water content alone. Cerebral ischemia produces a tissue edema consisting of a combination of intracellular and extracellular water accumulation within the tissue. The edema associated with ischemia has been assessed frequently using one or more noninvasive MR imaging techniques (1) in which the contrast has been based primarily on regional differences in proton density, spin-lattice (T 1 ), or spin-spin (T 2 ) proton relaxation times. Considering the fact that the majority of protons within the brain are found in water molecules, changes in intensity in T 1 -and T 2 -weighted images have often been interpreted to represent edema and have been used to monitor its progression under various pathologic states including cerebral ischemia and brain trauma (2). However, the actual changes in water content or the biophysical properties of the tissue that underlie many of the MR imaging changes remain poorly understood. Indeed, those studies measuring ontogenic or pathologic changes in tissue water content have reported a range of poor, mild, and good correlations between water content and MR variables such as T 2 and T 1 (3-12). The majority of such studies have not examined the correlation of the changes in MR relaxation variables with the evolution of the pathologic edema as a function of time, and even fewer have investigated directly the relationship between changes in MR variables and the edema resulting from a cerebral HI insult. Recently, we repor...