The majority of glioma patients experience declines in neurocognitive function (NCF), presumably due to tumor and treatment effects. We sought to understand the natural history of this decline in grade III and grade IV patients. A condensed battery of cognitive tests (HTLV, COWAT, GPT, TMT, and BTA) was administered at three time points: prior (T0) and after (T1) chemoradiation, and during adjuvant chemotherapy (T2). 31 patients were enrolled of which 25, 9 grade III and 16 grade IV, were analyzed. Although our N was too small for statistically significant results, we observed potentially meaningful clinical trends. Changes in HTLV, TMT, and COWAT scores among grade III patients displayed a pattern different from that of grade IV, with a steep decline seen after chemoradiation (T0-T1) followed by improvement several months later (T1-T2), despite ongoing chemotherapy. Grade IV patients, in contrast, showed minimal decline in scores (little change in score between T0-T1 and between T1-T2), perhaps highlighting that NCF is more impacted by disease rather than treatment. We tried to identify a subset of patients who seemed more susceptible to NCF decline. Examination of key clinical features showed that less than gross total resection and less than 4 year degree education level trended to associate with steeper NCF decline (only 40% of those experiencing steep declines in multiple domains had more than 4 yr degree, versus 60% of those with moderate or no decline). This pilot study highlights that assessing neurocognitive function routinely in clinical practice is feasible in a rural academic hospital. Based on patterns of changes in NCF, it appears that grade III and grade IV gliomas are distinct tumor subtypes with respect to NCF decline. Level of education may be a useful biomarker to identify those patients most at risk for neurocognitive decline after treatment.