Magnetic resonance imaging (MRI) techniques are capable of spatial and temporal resolution and within subject repeatability that far exceed other currently available methods of noninvasively investigating brain functioning. While functional MRI (fMRI) techniques have traditionally been used to investigate sensory, motor, and cognitive functions, they are also very attractive for investigating the effects of pharmacological agents in the brain. However, because pharmacological agents may interfere with the very mechanisms that give rise to the fMRI signal, careful attention to experimental design and data analysis issues must be exercised. Specifically, introducing a drug into the system could potentially alter the coupling of neural activity with regional cerebral blood flow and/or the extraction of oxygen from blood, or may cause local or global cardiovascular changes unrelated to neural activity. These concerns notwithstanding, careful attention to experimental detail and verification procedures promises to make pharmacological MRI a valuable tool for understanding the actions of drugs on the brain at many levels. The physiological and biophysical bases of blood oxygen level-dependent, arterial spin labeling, cerebral blood flow, cerebral blood volume, and spectroscopy acquisition methods are presented to illustrate where confounds may arise when a drug is added to the system. Selected studies combining administration of pharmacological agents with MRI techniques are critically reviewed to illustrate both the challenges inherent in these applications and some possible solutions. Included are studies of acute direct effects of drugs, effects of drugs on task-related activation, chronic effects of drugs, effects of drugs on cerebral metabolism, and variable effects of drugs in different populations.