The quantity of suitable habitat can be simply defined by the distribution and abundance of organisms, but it is the quality of the habitat that may have more direct demographic consequences. Habitat quality is best assessed by determining demographic rates and the performance of the focal species residing within a habitat. Using this framework, I used two demographic rates (settlement and survival) and two performance measures (growth and condition) of the giant kelpfish (Heterostichus rostratus) to evaluate whether eelgrass (Zostera marina) habitat quality varies spatially (i.e., site quality) within San Diego Bay, CA. Seven sites were selected that encompass the northern and central areas of the bay. Settlement was estimated bi-weekly from May-July 2009 with artificial seagrass units (ASUs) and standard monitoring units for the recruitment of fishes (SMURFs). Eelgrass habitat complexity (shoot height and density) was measured at each site, and three complexity treatments were replicated with ASUs in large mesocosms to estimate survival of giant kelpfish after exposure to kelp bass (Paralabrax clathratus) as predators. Growth and condition were derived from juvenile giant kelpfish collected from sites two months after the settlement period. Growth and condition were estimated from otolith daily growth increments and total lipid content, respectively. Settlement, growth, and condition were evaluated using a multivariate approach to provide an assessment of habitat quality among sites. Combined with habitat-dependent survival from predators in laboratory trials that reflected site-specific habitat attributes, sites were ranked along a gradient of habitat quality. The multivariate analysis of site quality based on demographic and performance variables was then overlaid with a similar analysis of environmental variables. Environmental variables that would serve as proxies for habitat quality were not concordant with the suite of demographic and performance variables used to differentiate site quality. Consequently, a multivariate approach using demographic rates and individual performance is the most appropriate method to evaluate habitat quality, whether among different habitats or among sites with the same habitat.