Knowledge of the demography and habitat requirements of the endangered Golden-cheeked Warbler (Setophaga chrysoparia) is needed for its recovery, including measures of productivity instead of reproductive indices. We report on breeding phenology and demography, calculate model-based estimates of nest survival and seasonal productivity and evaluate support for relationships with forest type, forest edge density, day of year, and year, and determine correspondence in these 2 measures of reproductive success. Males arrived in early March. Females laid the first egg of the first clutch in early April, made up to 5 nesting attempts, and completed nesting by mid-June. The most-supported nest survival model included day of year, proportion of juniper and juniper–oak forest within a 100-m radius of each nest, and the interactive effect of year and forest edge density. The most-supported seasonal productivity model included proportion of each forest type and the interactive effect of year and forest edge density. Seasonal productivity increased from 1.38 to 3.96 fledglings per territory and from 1.38 to 2.40 fledglings per territory across 0.00 to 0.87 and 0.00 to 1.00 proportion of juniper and proportion of juniper–oak forest, respectively. Seasonal productivity ranged from 1.86 to 3.12 fledglings per territory in 2010 and 2004, respectively (mean ± SD = 2.36 ± 0.37). Correlations between nest survival and seasonal productivity were strong when we controlled for the effect of year indicating demographic parameters other than nest survival, particularly renesting, double brooding, and polygyny, made an important contribution to actual seasonal productivity. The similarity in relationships of both measures of reproductive success with forest type and edge density and parallel findings for density with these habitat metrics reported in other studies provide strong rationale for protecting sites with high proportions of juniper and juniper–oak forest and less forest edge to further recovery efforts for the Golden-cheeked Warbler.