Providing accurate predictions of the thermodynamic properties of highly polar and hydrogen bonding compounds and their mixtures is challenging from a theoretical perspective. The combination of an equation of state (EoS) based on the statistical associating fluid theory (SAFT) with a group contribution (GC) methodology offers both accuracy and predictive capability for the thermodynamic properties of mixtures. In our current work, the SAFT-γ Mie equation of state is used to capture the underlying complexity of systems in which specific interactions (e.g., hydrogen bonding, dipolar interactions, chemical association) play an important role, by incorporating highly versatile association-site schemes to model mixtures in which unlike induced association interactions occur; this is done by assigning to the functional groups a number of association sites that are inactive in the pure fluid, but become active in certain mixtures. We refer to this type of association mechanism as "unlike induced" association and to the sites involved in this interaction as "unlike induced" association sites. The concept of unlike induced association sites is applied here to develop reliable SAFT-γ Mie group contribution models to describe the properties of acetone, alkyl carboxylic acids, and their mixtures with water and n-alkanes. The parameter table of available SAFT-γ Mie models is expanded to incorporate the corresponding group interaction parameters for acetone, which is treated as a molecular group, the carboxyl group COOH, and their unlike interaction group parameters with water, and the methyl CH 3 , methanediyl CH 2 , and methanetriyl CH alkyl groups. In particular, one unlike induced site is used with the acetone model to mediate hydrogen-bonding of the acetone oxygen in mixtures containing hydrogen-bond donors, and two pairs of unlike induced sites are included on the COOH group to mediate hydrogen-bond formation in mixtures of carboxylic acids and hydrogen-bond donors. The models developed allow for the successful description of the complex fluid-phase behaviour of the relevant binary and ternary mixtures, including accurate predictions of systems which have not been used to determine the model parameters.