In the present study, we have correlated plasma TDP-43 levels, as measured by ELISA, with the presence of TDP-43 pathological changes in the brains of 28 patients with frontotemporal lobar degeneration (FTLD) (14 with FTLD-TDP and 14 with FTLD-tau) and 24 patients with pathologically confirmed AD (8 with, and 16 without, TDP-43 pathological changes). Western blotting revealed full-length TDP-43, including a phosphorylated form, and a phosphorylated C-terminal fragment, in all samples examined. Both ELISA and immunohistochemistry were performed using phospho-dependent and phospho-independent TDP-43 antibodies for detection of phosphorylated and total TDP-43, respectively. Over all 52 cases, plasma levels of TDP-43, and scores of brain TDP-43 pathology, determined using TDP-43 phospho-dependent antibody correlated with the equivalent measure determined using the TDP phospho-independent antibody. In FTLD, but not AD, TDP-43 plasma levels correlated significantly with the pathology score when using the TDP-43 phospho-dependent antibody, but a similar correlation was not seen in either FTLD or AD using the TDP-43 phospho-independent antibody. With the TDP-43 phospho-independent antibody, there were no significant differences in median plasma TDP-43 levels between FTLD, or AD, patients with or without TDP-43 pathology. Using TDP-43 phospho-dependent antibody, median plasma TDP-43 levels were greater in patients with, than in those without, TDP-43 pathology for FTLD patients, though not significantly so, but not for AD patients. Present assays for TDP-43 do not differentiate between FTLD, or AD, patients with or without TDP-43 pathological changes in their brains. However, the levels of phosphorylated TDP-43 in plasma do correlate with the extent of TDP-43 brain pathology in FTLD, and therefore might be a useful surrogate marker for tracking changes in TDP-43 brain pathology during the course of this disease.