Critical interactions between genetic and environmental factors -- among which stress is one of the most potent non-genomic factors -- are involved in the development of mood disorders. Intensive work during the past decade has led to the proposal of the network hypothesis of depression [Castren E: Nat Rev Neurosci 2005;6:241-246]. In contrast to the earlier chemical hypothesis of depression that emphasized neurochemical imbalance as the cause of depression, the network hypothesis proposes that problems in information processing within relevant neural networks might underlie mood disorders. Clinical and preclinical evidence supporting this hypothesis are mainly based on observations from depressed patients and animal stress models indicating atrophy (with basic research pointing at structural remodeling and decreased neurogenesis as underlying mechanisms) and malfunctioning of the hippocampus and prefrontal cortex, as well as the ability of antidepressant treatments to have the opposite effects. A great research effort is devoted to identify the molecular mechanisms that are responsible for the network effects of depression and antidepressant actions, with a great deal of evidence pointing at a key role of neurotrophins (notably the brain-derived neurotrophic factor) and other growth factors. In this review, we present evidence that implicates alterations in the levels of the neural cell adhesion molecules of the immunoglobulin superfamily, NCAM and L1, among the mechanisms contributing to stress-related mood disorders and, potentially, in antidepressant action