Goal-Oriented Requirements Engineering (GORE) offers proved means to decompose technical and non-technical requirements into well-defined entities (goals) and reason about the alternatives to meet them. Hence, it has been used as a means to model and reason about the systems’ ability to adapt to changes in dynamic environments. PistarGODA-MDP is a goal-oriented framework to model self-adaptive systems (SAS) under different classes of uncertainty, namely (i) system itself, (ii) system’s goals, and (iii) environment. The framework augments a Contextual Goal Model (CGM) with new annotations to represent uncertainties while acknowledging the nondeterminism brought by them. Then, it automatically generates: (i) a Markov Decision Process (MDP) model in PRISM language, and (ii) reliability and cost parametric formulae of the corresponding system. The former is an input used by probabilistic model checking activity to support system analysis and verification at design time. The latter is useful for the guidance of SAS adaptation policies at design time, and for runtime verification of system reliability and cost statuses.