Cognitive impairment associated with schizophrenia (CIAS) and related deficits in learning (plasticity) are among the leading causes of disability in schizophrenia. Despite this, there are no Food and Drug Administration-approved treatments for CIAS, and the development of treatments has been limited by numerous phase 2/3 failures of compounds that showed initial promise in small-scale studies. NMDA-type glutamate receptors (NMDARs) have been proposed to play an important role in schizophrenia; moreover, the NMDAR has a well-characterized role in cognition, learning, and neuroplasticity. We review previously published clinical trials in CIAS that focused on NMDAR modulator treatments, focusing on published and recent developments of the use of novel NMDAR-modulating treatments for CIAS both alone and combined with plasticity/learning paradigms to enhance learning. We use this discussion of previous studies to highlight the importance of incorporating pharmacodynamic target engagement biomarkers early in treatment development, which can help predict which compounds will succeed or fail in phase 3. A range of direct and indirect NMDAR modulators are covered, including D-serine, D-cycloserine, memantine, and glycine and first-generation glycine transport inhibitors (e.g., sarcosine and bitopertin), as well as recent positive studies of iclepertin, a novel glycine transport inhibitor, and luvadaxistat, a D-amino acid oxidase inhibitor that increases brain D-serine levels, and indirect noninvasive brain stimulation NMDAR-modulating treatments. Several examples of successful use of pharmacodynamic target engagement biomarkers for dose/drug discovery are emphasized, including the mismatch negativity, auditory steady state, and time-frequency event-related potential approaches.