Probabilistic fuzzy rules differ from traditional fuzzy if-then rules, which typically do not account for probabilistic aspects of the output. In standard fuzzy systems, rules are usually formulated purely in terms of fuzzy logic, without explicit consideration of the underlying probability distributions of the output variables. The probabilistic fuzzy rules extend the traditional fuzzy rules by incorporating a probabilistic dimension into the consequent, providing a richer framework for modeling and reasoning under uncertainty. In this presentation, we overview some selected approaches to probabilistic fuzzy modeling and provide some new directions in this field.