Long-term potentiation (LTP) can only be induced at a synapse up to a fixed maximum extent; the magnitude of this ceiling has been found to be modulated by both endogenous and exogenous factors. The effect of the LTP ceiling on network-level categorization performance in a simulated layer of cortical neurons is investigated. The simulation is presented with random-dot stimuli and the similarity of its outputs is used to determine how it classifies its inputs. It was found that a higher ceiling leads to wider category formation, thus suggesting additional computer simulations and human experiments to test the connection between this LTP parameter and categorization performance.