Sensory prostheses can restore aspects of natural sensation by delivering electrical current directly into sensory circuits. An effective sensory prosthetic should be capable of generating reliable real-time perceptual signals for hours each day over many years. However, we still know little regarding the stability of percepts produced by electrical microstimulation of cerebral sensory cortex when stimulation is delivered repeatedly over long periods. Developing methods that yield highly sensitive and reliable assessments of a subject's sensitivity to stimulation is important for developing prosthetic devices that can mimic the constant stream of information inherent in daily experience. Here, we trained rhesus monkeys to report electrical microstimulation of their primary visual cortex (V1) and measured how repeated stimulation affected the minimal electrical current needed to generate a percept (behavioral detection threshold). Using adaptive staircase procedures with a two-alternative forced-choice (2AFC) detection task, we obtained highly reliable detection threshold measures with as few as 100 trials. Using either chronically implanted or acutely inserted microelectrodes, we found that repeated electrical microstimulation elevated detection thresholds, with effects persisting between daily testing sessions. Our results demonstrate task designs that can support rapid and reliable measurements of detection thresholds, and point to the need for validation that detection thresholds in targeted structures will be sufficiently stable in the face of the amount of chronic stimulation that will be required for effective sensory prosthetics.