This paper describes how a design repository can be used as a concept generation tool by drawing upon archived function-based design knowledge. Modern design methodologies include several types of activities to formally generate design concepts. Typical concept generation methods range from open-ended creative brainstorming activities to quantitative function-component analysis. A combination of two such methods—the chi-matrix and morphological matrix techniques—is the basis for this work. Building on existing functionality of the design repository, desired product functions can be specified in a search of stored design knowledge, returning a morphological matrix of artifacts solving the specified functions. Such a search is termed a morphological search. The repository morphological search feature is evaluated against concepts generated in a previous original design project. Results of the morphological search return are then compared to ten of the original concept variants generated during the design project. This comparison shows that 89% of the specified subfunctions return results and that, on average, 77% of the components used in the hand-generated concepts can be derived by using the morphological search feature.