In previous experiments, a causal relationship between sodium influx and secretion of nerve growth factor (NGF) was deduced from the observation that the sodium substitute N-methyl-D-glucamine (NMDG) abolished any activity-mediated NGF secretion that depends on intact internal calcium stores. However, all available experimental evidence speaks against sodium-mediated calcium mobilization from these stores under physiological conditions. We now report that rapid sodium influx initiated by monensin or ouabain did not induce brain-derived neurotrophic factor (BDNF) secretion from either native hippocampal slices or BDNF-transduced hippocampal neuronal cultures. Additionally, we found marked differences between the replacement of sodium by NMDG and sucrose on the one hand, and choline and lithium on the other. Replacement of 100% (and as little as 10%) sodium by NMDG or sucrose not only blocked the activity-mediated neurotrophin secretion, but itself led to a rapid and substantial increase of neurotrophin secretion. In contrast, the replacement of sodium (10% and 100%) by lithium and choline did not result in a release of neurotrophins, and only 100% replacement blocked the activity-mediated neurotrophin secretion. We conclude that the blocking effects of NMDG and sucrose on neurotrophin secretion do not reflect the sodium replacement, but instead represent an independent blocking effect. These differences were also reflected in part by electrophysiological investigations in individually patched hippocampal neurons. The importance of the present observations lies not only in the reevaluation of the involvement of sodium in activity-dependent neurotrophin secretion, but also in the demonstration that sodium replacement may initiate 'side effects' that are unrelated to sodium replacement.