We investigated the dose-response relationship of protection by creatine against ischemic damage, and we asked whether or not such protection may be observed in invertebrate neurons that might provide a simpler experimental model. Rat isolated pyramidal neurons from the CA3 region of hippocampus subjected to ischemia ("in vitro ischemia") showed anoxic depolarization (AD) after 3-7 min of incubation in anoxic medium. Membrane potential (MP) was reduced 25-78% from preanoxic value. Inward current was decreased by an average 49%. Supplementation with creatine protected against these changes, with 1 mM being the minimal effective concentration, 2 mM providing a near-maximal protection, a maximal effect being obtained with 5 mM creatine. No additional protection was provided by up to 20 mM creatine. Isolated giant neurons of Lymnaea stagnalis showed a similar response to in vitro ischemia. However, a clear seasonal dependence of sensitivity of these cells was detected. In cells obtained during summertime (May-August), AD latency ranged from 3 to 10 min; during wintertime (December-March), this response did not occur even after 25-50 min. The addition of creatine to the medium did not cause changes in AD latency, probably because these neurons rely on a phosphoarginine/arginine energy system. However, treatment of the cells, harvested during summertime, with 2 mM arginine did provide clear protection against anoxic-aglycaemic changes. Summing up, besides confirming previous findings on creatine protection in mammalian neurons, we (1) better characterized their dose-response relationship and extended the findings to the CA3 region and to isolated neurons, (2) found that invertebrate neurons are not protected by creatine but by arginine supplementation and (3) reported a novel mechanism of seasonal dependence in sensitivity of in vitro ischemia by invertebrate neurons.