In the study I analyze the conflicting aspects of project portfolio evolution in a firm. The evolutionary principles of variation, selection and retention are applied to the management of new product development projects. Managers select projects for prioritization. A selection rule is the prioritization rule. In biology, living creatures develop specific features for adaptation as a result of selection rules. However, the selection of specific adaptive features carries along the retention of other, even unforeseen non-adaptive features. Drawing on the evolutionary principles forwarded by Darwin I examine how they manifest in the project portfolio. I define this non-adaptive mechanism as co-selection. By analogy, in portfolio management, if the selection rule for project priority is high revenue and feasibility to global access, other features also survive when the selection rule relating to the prioritization of projects is applied. The evolution of the new product development project portfolio in the case firm displays conflicting trends in the emerging project portfolio over time. Managers pursue prioritization to decrease product development times. But, alas, in the project portfolio the prioritized projects age to a greater degree than non-prioritized projects. Managers prioritize the projects held by the focal business unit more often than those of other business units. However, ultimately the focal business unit has less than a due share of prioritized projects in the portfolio. The results of this study question the applicability of optimizing models in R&D portfolio management in the presence of co-selection. The project portfolio management literature does not provide a mechanism to account for this type of portfolio development. Co-selection provides a mechanism that explains the observed evolution. The study contributes to the conceptualization of the notion of co-selection. The study also provides empirical evidence on co-selection, a non-adaptive evolutionary mechanism to modify R&D project portfolio outcome. The findings give a better understanding of portfolio management of R&D driven new product development projects.