Pyramidal training is an effective model for disseminating behavior analytic skills. However, pyramidal training in research is often conducted in controlled university settings. Further, research that has evaluated the effectiveness of pyramidal training in classroom settings (see Pence et al. 2014) often focuses on improving the use of one procedure (e.g., functional analysis) over a brief period. We conducted this study to evaluate the generalized effectiveness of behavioral skills training within a pyramidal model to improve teacher training skills across five procedures in the classroom setting over a nine-month period. We used behavioral skills training (BST) to train teachers (Tier 1) to implement BST to train teaching assistants (Tier 2) to implement Applied Behavior Analytic (ABA) procedures stimulus-stimulus pairing, multiple stimulus without replacement (MSWO) preference assessment, mand training, discrete trial teaching (DTT), and graphing data. Pyramidal training was effective in increasing teachers' procedural integrity of BST, and the social validity of the model was apparent in that teacher assistants' procedural integrity of target ABA procedures increased after Tier 2 training. Teachers required periodic feedback to maintain training skills, train novel procedures, and novel staff. Thus, pyramidal BST was effective to teach new skills, but required ongoing monitoring and feedback to ensure maintenance and generality of training skills.