This paper constitutes an innovative attempt to analyse the risks and negative phenomena dependencies within a project portfolio. Based on the available literature, the risks and negative phenomena (that is, the problems with the availability of resources, interpersonal conflicts, irregularities in the portfolio balance, etc.) specific to a project portfolio were identified. Theoretical constructs were then used to connect the identified risks with the negative phenomena. Structural equations were used to confirm the existence and quality of these constructs, as well as models describing connections between phenomena. The determination of the structural equations also provided a setting in which statistical methods (χ 2 , RMSEA and CFI) could be used to investigate the level of fit of the constructs and models to the empirical data.