Autism spectrum disorder is characterized by communication and/or social deficits with restricted and repetitive behaviors. Treating autism is very costly, both financially and emotionally. Early Intensive Behavioral Intervention (EIBI) has been shown to decrease the symptomology for those with ASD; although, we cannot predict who will benefit from EIBI at this time. Discrete trial data were used for 15 students enrolled in EIBI in addition to developmentally appropriate social training. Individual student trajectories through time spent in therapy were analyzed using ARIMA modeling, and predictor variables of post-treatment gains were also explored. Time spent mastering basic skill programs significantly predicted post-Mullen subscale gains. Joint Attention is also a significant predictor. Also, error variance in the most complex of the five basic skill programs, One-step Directions, was a significant predictor of post-therapy gains. These potential tailoring variables to assess non-responders early in therapy will hopefully prove useful in individualizing treatment for children with ASD.