The therapeutic results of systemic administration of pralidoxime (2-PAM) in the treatment of poisoning with organophosphate-type cholinesterase inhibitors are disappointing. It has been hypothesized that this is due to poor entry of 2-PAM into the brain. To test if survival rates can be improved by direct administration of 2-PAM into the cerebrospinal fluid (CSF), the effect of intrathecal 2-PAM injections upon mortality after paraoxon intoxication was examined. Eight groups of rats (n=30 each) were examined, all of which received paraoxon (1 micromol=272 microg, 3 micromol=816 microg, or 5 micromol=1.36 mg) intraperitoneally (i.p.). One group received no further treatment; the other groups were given 50 micromol (=8.63 mg) 2-PAM i.p., 5 micromol (=863 microg) 2-PAM intrathecally and pentobarbital/lidocaine in various combinations. Results were compared with the no treatment group and the control groups that did not receive any paraoxon injections, but were given intrathecal injections of saline or 2-PAM. The relative risk of death was estimated by Cox survival analysis. Mortality was lowest after treatment with a combination of both i.p. and intrathecal 2-PAM plus pentobarbital, and with 2-PAM i.p. alone plus pentobarbital. Both treatments were significantly better than 2-PAM i.p. alone (p<or=0.0001). Mortality of intrathecal 2-PAM plus either pentobarbital or lidocaine was in the same order of magnitude as 2-PAM i.p. without pentobarbital/lidocaine, which was significantly (p<or=0.05) lower than that of the no treatment group and of the groups that had only been given pentobarbital or lidocaine. Our results indicate that i.p. injections of 2-PAM insufficiently protect the CNS against the effect of paraoxon. Supplementary injection of pentobarbital, presumably by its anticonvulsive action, has a superior efficacy compared to 2-PAM i.p. alone, irrespective of additional intrathecal 2-PAM administration.