In previous studies, we found large rises in extracellular potassium activity ([K+]e) and falls in extracellular Ca2+ activity ([Ca2+]e) in injured spinal cords. [K+]e recovered rapidly at the impact site within 2 h but ischemia onset at 2-3 h paradoxically did not cause further [K+]e rises. [Ca2+]e, in contrast, remained depressed for long periods of time, suggesting either an effective block of Ca diffusion to the injury site or a deep Ca2+ sink at the impact site. To resolve questions raised by the [K+]e and [Ca2+]e recovery patterns, we used atomic absorption spectroscopy to measure spatial distributions of tissue concentrations of K ([K]t) and Ca ([Ca]t) in cat spinal cords injured by a standardized contusion, compared with uninjured controls. At the impact site, [K]t fell to 51% and 35% of control at 1 and 3 h. The K content of cord surrounding the impact site did not change significantly at 1 h, but K gains in surrounding cord at 3 h approximated K losses from the impact site. The K results indicate that contusion disrupts greater than 80% of cells at the impact site with K loss to adjacent cord, blood, and cerebrospinal fluid. Such losses may explain why subsequent ischemia at the impact site did not cause [K+]e rises. [Ca]t at the impact site increased to 37% and 59% above control at 1 and 3 h. The Ca gain at the impact site exceeded the amount of free extracellular Ca2+ available before injury within 2 cm of the impact site. At 1 h, the Ca lost in cord surrounding the impact site approximated the Ca gain at the impact site. These findings indicate that Ca accumulated at the impact site comes largely from surrounding cord. We propose that Ca sequestration by inorganic phosphates causes a deep Ca sink at the impact site.