The independent effects of extranasal-only carriage, carriage at multiple bodily sites, or the bacterial load of colonizing Staphylococcus aureus (SA) on the risk of developing SA surgical site infections and postoperative bloodstream infections (SA SSI/BSIs) are unclear. We aimed to quantify these effects in this large prospective cohort study. Surgical patients aged 18 years or older were screened for SA carriage in the nose, throat, or perineum within 30 days before surgery. SA carriers and noncarriers were enrolled in a prospective cohort study in a 2:1 ratio. Weighted multivariable Cox proportional hazard models were used to assess the independent associations between different measures of SA carriage and occurrence of SA SSI/BSI within 90 days after surgery. We enrolled 5004 patients in the study cohort; 3369 (67.3%) were SA carriers. 100 SA SSI/BSI events occurred during follow-up, and 86 (86%) of these events occurred in SA carriers. The number of colonized bodily sites (adjusted hazard ratio [aHR], 3.5-8.5) and an increasing SA bacterial load in the nose (aHR, 1.8-3.4) were associated with increased SA SSI/BSI risk. However, extranasal-only carriage was not independently associated with SA SSI/BSI (aHR, 1.5; 95% CI, 0.9-2.5). Nasal SA carriage was associated with an increased risk of SA SSI/BSI and accounted for the majority of SA infections. Higher bacterial load, as well as SA colonization at multiple bodily sites, further increased this risk.