Abstract Pseudomonas aeruginosa populations evolving in cystic fibrosis (CF) lungs, animal infection models, natural environments or in vitro undergo extensive genetic adaption and diversification. A common mutational target is the quorum sensing (QS) regulon, a three-unit regulatory system that controls the expression of a suite of virulence factors and secreted public goods. Three scenarios have been advocated to explain selection for QS mutants, which include (I) disuse of the regulon, (II) cheating on public goods, or (III) modulation of the network. Here, we test these scenarios by examining a set of 61 QS mutants from an experimental evolution study. We observed non-synonymous mutations in all three QS systems – Las, Rhl and PQS. Most Las mutants carried large deletions, resulting in loss of QS function, and the inability to produce QS-regulated traits (scenario I or II). Conversely, phenotypic and gene expression analyses of Rhl mutants support network modulation (scenario III), as these mutants overexpressed the Las and Rhl regulators and showed an altered QS-regulated trait production portfolio. PQS mutants also showed patterns of network modulation (scenario III), spurring strain diversification and phenotypic trade-offs, where the upregulation of certain QS traits is associated with the downregulation of others. Overall, our results indicate that mutations in different QS systems lead to diverging effects on the social portfolio of bacterial populations. These mutations might not only affect the plasticity and diversity of evolved populations but could also impact bacterial fitness and virulence in infections. Importance Pseudomonas aeruginosa uses quorum sensing (QS), a three-unit multi-layered network, to coordinate expression of traits for growth and virulence in the context of infections. Despite its importance for bacterial fitness, the QS regulon appears to be a common mutational target during long-term adaptation of P. aeruginosa in the host, natural environments and experimental evolutions. This raises the questions why such an important regulatory system is under selection and how mutations change the portfolio of QS-regulated traits. Here, we examine a set of 61 naturally evolved mutants to address these questions. We found that mutations involving the master regulator, LasR, resulted in an almost complete breakdown of QS, while mutations in RhlR and PqsR resulted in modulations of the QS regulon, where both the QS regulon structure and the QS-regulated trait portfolio changed. Our work reveals that natural selection drives diversification in QS activity patterns in evolving populations.