Selection for improved sow lifetime reproductive productivity is of growing importance with recent increases in sow culling rates. Minimal response occurs with traditional selection due to low heritability and delayed expression of the trait; thus, marker-assisted selection may be a more effective alternative. In order to assess the relationship between markers and reproductive longevity, 967 gilts from two maternal crossbred lines were evaluated for: age at puberty (AP), measures of rate and composition of growth, and lifetime production (LP) defined as total born through four parities. After random culling to accommodate available farrowing space, females were culled only for death, unsoundness, or reproductive failure. Gilts that reached puberty by 240 days of age (n=851) were designated as breeders and genotyped for 62,183 SNPs. Genome-wide association studies employing Bayes C and B approaches were performed for AP, LP and several other pre-breeding growth traits, with AP displaying the largest proportion of phenotypic variance explained by SNP effects (32%). Eleven 1 Mb regions that harbor rich clusters of SNPs associated with AP explained the largest proportion of phenotypic variation in AP and several of these regions appeared to influence both AP and LP. Functional annotation, enrichment, and pathway analyses were performed in these regions to identify potential sources of phenotypic variation. Two regions were found to coincide with previously reported QTLs associated with AP. The incorporation of identified markers into a breeding program can potentially be used to predict age at puberty and to improve sow lifetime reproductive productivity.