Both visceral obesity and depression are associated with impaired health and excess mortality, possibly through overlapping pathophysiological mechanisms like adipose tissue derived inflammatory markers. These results, however, are primarily based on population-based surveys, often restricted to a young population and depression severity scales instead of patients with established diagnosis of depressive disorder. We examined the relation between waist circumference and late-life depression using the baseline data of the Netherlands Study of Depression in Older people (NESDO). Psychopathology has been assessed with Composite International Diagnostic Interview version 2.1. Adjusted for age, sex, education, lifestyle (smoking, alcohol, physical activity), drug use, cognition and chronic diseases as well as adjusted for body mass index (BMI), analysis of covariance showed that depressed older patients (n = 376) had a significantly lower waist circumference (WC) compared to their non-depressed comparisons (n = 130): estimated marginal mean (SE) = 93.9 (0.5) versus 97.8 (0.8) cm (F = 15.9; df = 1467; p < .001). Multiple linear regression analyses within the depressed group showed that both, depression severity (Inventory of Depressive Symptoms) as well as duration-related depression characteristics (age of onset, duration of illness, life-time comorbid dysthymia), were associated with the WC. Only the severity of depressive symptoms remained significant after further adjusted for the BMI. Interestingly, a recently discovered adipokine, Neutrophil Gelatinase-Associated Lipocalin (NGAL), was associated with late-life depression, but only in the subgroup of patients with a pathologically increased WC.