Current theory about online sexual predation suggests that predators engage in a grooming process with their victims. Grooming consists of three main stages: friendship and relationship forming, risk assessment and exclusivity, and sexual. This thesis examines whether predators utilized different patterns of language use across each of the stages that can be identified through computerized text analysis. The transcripts of 43 convicted predators were spliced into three equal sections according to word count and a computerized text analysis was performed to look at the different types of language that are expected in each stage based on grooming. The results reveal that predator language differs significantly throughout the three stages of grooming, and that the grooming stages can predict language patterns that are used most frequently in each stage. The theoretical and practical implications of these findings are discussed in terms of grooming theory and the use of computerized text analysis to identify predators and educate youth.