Despite kratom’s growing popularity in the United States and claims of its link to opioid use, research has yet to fully delineate the factors that lead to its use. Recent research has primarily focused on kratom user-profiles and prevalence in the U.S. This current study explores patterns and associations behind kratom use within a nationally representative sample collected by the Substance Abuse and Mental Health Services Administration (SAMHSA). We apply a machine-learning technique to consider the possibility of predicting kratom use among a large sample and whether kratom use plays a role in opioid cessation. The results show that it was not possible to predict kratom use among the general dataset, even when isolating opioid misusers from the general sample. Further, the results show that kratom use did not play a role in opioid cessation. While it was not possible to predict kratom use, notable predictors were highlighted during this study that has not been previously explored. The findings hold important implications for policy and future research related to kratom.