This paper provides a tutorial on using generalized boosted models (GBM) to estimate propensity scores for multiple treatments. Propensity scores are used to control for pretreatment imbalances in observational studies. While tools exist for two treatments, guidance for three or more treatments is limited. The paper aims to provide step-by-step guidance for implementing propensity score weighting for multiple treatments and proposes GBM for estimating propensity score weights. It defines causal quantities of interest and derives weighted estimators for these quantities. The paper presents a detailed plan for using GBM to estimate propensity scores and weights, and provides tools for assessing balance and overlap of pretreatment variables. A case study on three treatment programs for adolescent substance abuse demonstrates the methods. The paper discusses causal effects with multiple treatments, including average treatment effects (ATE) and average treatment effects among the treated (ATT). It describes the method for estimating multiple treatment effects using propensity score weights, including the assumptions and conditions for inverse probability of treatment weighting (IPTW). The paper also discusses estimating multiple propensity scores using multinomial logistic regression and GBM, and provides methods for assessing balance. The paper concludes that GBM can mitigate challenges associated with multinomial regression and provides a practical approach for estimating propensity scores and weights in multiple treatment settings.This paper provides a tutorial on using generalized boosted models (GBM) to estimate propensity scores for multiple treatments. Propensity scores are used to control for pretreatment imbalances in observational studies. While tools exist for two treatments, guidance for three or more treatments is limited. The paper aims to provide step-by-step guidance for implementing propensity score weighting for multiple treatments and proposes GBM for estimating propensity score weights. It defines causal quantities of interest and derives weighted estimators for these quantities. The paper presents a detailed plan for using GBM to estimate propensity scores and weights, and provides tools for assessing balance and overlap of pretreatment variables. A case study on three treatment programs for adolescent substance abuse demonstrates the methods. The paper discusses causal effects with multiple treatments, including average treatment effects (ATE) and average treatment effects among the treated (ATT). It describes the method for estimating multiple treatment effects using propensity score weights, including the assumptions and conditions for inverse probability of treatment weighting (IPTW). The paper also discusses estimating multiple propensity scores using multinomial logistic regression and GBM, and provides methods for assessing balance. The paper concludes that GBM can mitigate challenges associated with multinomial regression and provides a practical approach for estimating propensity scores and weights in multiple treatment settings.