In the coronavirus efficacy (COVE) phase 3 efficacy trial of the mRNA-1273 vaccine, IgG binding antibody (bAb) concentration against Spike (BA.1 strain) and neutralizing antibody (nAb) titer against Spike (BA.1 strain) pseudovirus were assessed as correlates of risk of Omicron COVID-19 and as correlates of relative boost efficacy in per-protocol recipients of a third (booster) dose. Markers were measured on the day of the boost (BD1) and 28 days later (BD29). For SARS-CoV-2 naive individuals, BD29 Spike IgG-BA.1 strain bAbs and BD29 BA.1-strain nAbs inversely correlated with Omicron COVID-19: hazard ratio (HR) per 10-fold marker increase [95% confidence interval (CI)] = 0.16 (0.03, 0.79); P=0.024 and 0.31 (0.10, 0.96); P = 0.042, respectively. These markers also inversely correlated with Omicron COVID-19 in non-naive individuals: HR = 0.15 (0.04, 0.63); P = 0.009 and 0.28 (0.07, 1.08); P = 0.06, trend. Fold-rise in markers from BD1 to BD29 had similarly strong inverse correlations. For SARS-CoV-2 naive individuals, overall booster relative (three-dose vs two-dose) efficacy was 46% (95% CI: 20%, 64%) and correlated with BA.1 strain nAb titer at exposure. At 56, 251, and 891 arbitrary units (AU)/ml (10th, 50th, and 90th percentile), the booster relative efficacies were -8% (95% CI: -126%, 48%), 50% (25%, 67%), and 74% (49%, 87%), respectively. Similar relationships were observed for Spike IgG-BA.1 strain bAbs and for the markers measured at BD29. The performance of bAb and nAb markers as correlates of protection against Omicron COVID-19 supports their continued use as surrogate endpoints for mRNA vaccination against Omicron COVID-19.