Many individuals experience significant positive psychological change after ingesting psychedelic agents such as psilocybin, LSD, and ayahuasca, but some do not. Such changes appear to be mediated by definable and measurable mystical experiences (MEs) during the psychedelic session, which have been shown in previous research (e.g., Griffiths, Richards, McCann, & Jesse, 2006) to be related to long-term positive personal change. Individual responses to psychedelic agents are characteristically difficult to predict, however, necessitating improved identification of predictors to (a) reduce the possibility of significant challenging experiences and sequelae and (b) increase the possibility of MEs and long-term beneficial outcomes. Our study utilized crowdsourced retrospective survey data from 143 Amazon Mechanical Turk participants who self-administered psilocybin in the past year to identify factors related to MEs and/or challenging experiences. The best performing regression model explained 66% of the variance in ME scores on the 30-item Mystical Experience Questionnaire (MacLean, Leoutsakos, Johnson, & Griffiths, 2012) and 56% of the variance in scores on the dread subscale of the Sacred Emotions Scale (Burdzy, 2014). A state of surrender at the start of the psilocybin session most strongly explained MEs, and a state of preoccupation most strongly explained challenging experiences. The trait absorption was a strong secondary predictor in both models, along with smaller predictors. Additionally, the occurrence of MEs during the psilocybin session explained long-term positive change. Based on these findings, the presented models appear to provide a powerful framework for explaining the immediate quality and long-term consequences of psychedelic experiences. Limitations as well as implications for therapeutic applications are discussed.