The c-Jun N-terminal kinases (JNKs) are stress-activated serine-threonine kinases that have recently been linked to various neurological disorders. We previously described a patient with intellectual disability (ID) and seizures (Patient 1), carrying a de novo chromosome translocation affecting the CNS-expressed MAPK10/JNK3 gene. Here, we describe a second ID patient (Patient 2) with a similar translocation that likewise truncates MAPK10/JNK3, highlighting a role for JNK3 in human brain development. We have pinpointed the breakpoint in Patient 2, which is just distal to that in Patient 1. In both patients, the rearrangement resulted in a predicted protein interrupted towards the C-terminal end of the kinase domain. We demonstrate that these truncated proteins, although capable of weak interaction with various known JNK scaffolds, are not capable of phosphorylating the classical JNK target c-Jun in vitro, which suggests that the patient phenotype potentially arises from partial loss of JNK3 function. We next investigated JNK3-binding partners to further explore potential disease mechanisms. We identified PSD-95, SAP102 and SHANK3 as novel interaction partners for JNK3, and we demonstrate that JNK3 and PSD-95 exhibit partially overlapping expression at synaptic sites in cultured hippocampal neurons. Moreover, JNK3, like JNK1, is capable of phosphorylating PSD-95 in vitro, whereas disease-associated mutant JNK3 proteins do not. We conclude that reduced JNK3 activity has potentially deleterious effects on neuronal function via altered regulation of a set of post-synaptic proteins.