Inferotemporal cortex plays critical role in analysis and storage of visual information. Selective neurons in this region of the primate's brain are known to modulate their activity in response to specific patterns, including complex shapes, objects, and faces. While many studies have examined neuronal selectivity in this region of the brain, little is known about the maintenance of such selectivity over a period of days and weeks. Traditional recording techniques have provided only indirect information on this issue, and, given the suspected malleability of selective responses in these areas, it would be of great value to learn of the relative permanence of selective representations in this area. Recent advances in implantable electrodes have made it possible to record chronically from isolated single units for periods of days and weeks. We tested monkeys on a simple fixation task, presenting a large number of complex patterns and images while monitoring action potentials with the 64 implanted microwires. Using methods that have been traditionally used to evaluate the stability of neurons from day to day, namely comparing amplitude-normalized spike waveforms and interspike interval distributions, we found that the parameters of our measured neurons were often statistically indistinguishable from day to day. Units showing unchanging normalized waveforms and interspike interval distributions were considered for further analysis. We found that each neuron's response amplitude and dynamics, including its latency, transience, and excitation/inhibition, differs markedly for the different stimuli, but is nonetheless similar across recording sessions. The stability in wave forms, combined with maintained complex selectivity from session to session, provided strong evidence that the same neurons were being successfully monitored over days and weeks. Across the population, we found that neurons retained their pattern of selectivity from day to day. This was evaluated by applying a correlationbased similarity index (CSI), which takes into account both the magnitude and temporal response to each of the 65 stimuli that were presented. We found that the CSI was much higher for neurons identified as being the same across sessions (median 0.67, n=45), than for neurons collected from different neurons between sessions (median 0.18, n=883). For different neurons collected on the same day from the same electrode, the CSI (median=0.31) was much lower than the CSI the same neuron monitored across days. The results suggest that individual neurons have remarkably specific and fixed roles in analysis of complex stimuli over a period of days.