Modern symptom validity tests (SVTs) use empirical cutoffs for decision making. However, limits to the applicability of these cutoffs may arise when severe cognitive symptoms are present. The purpose of the studies presented here was to explore these limits of applicability. In Experiment 1, a group of 24 bona fide neurological patients without clinically obvious cognitive symptoms was compared to a group of 24 patients with rather severe symptoms. A comprehensive test battery was employed, which included four SVTs (the Test of Memory Malingering, TOMM, the Word Memory Test, WMT, the Bremer Symptomvalidierung, BSV, and the Amsterdam Short-Term Memory Test, ASTM). In Experiment 2, a group of 20 patients with mild Alzheimer's disease was compared to 14 healthy controls. Results of both studies showed that cognitive impairment may significantly interfere with SVT performance. Correlation analyses revealed dissimilar relationships between SVTs and neuropsychological test measures. Whereas TOMM and WMT correlated mainly with tests of declarative memory, the BSV correlated with tests of attention, and ASTM correlated with tests of working memory. Intercorrelations between symptom validity measures were relatively low. The published cutoffs of the TOMM would be suitable for estimating effort in patients with Mini-Mental State Examination scores of 24 or above. More research will be necessary to investigate how performance in SVTs is related to cognitive functioning in populations of severely impaired patients.