Multiple binary outcomes occur quite frequently in oral health research, as well as other areas of health care research. When there is interest in comparing whether covariates influence one outcome more than another, statistical methods that adjust for the correlation that may exist between outcomes are warranted. Available software is limited to the extent that some pre-processing of the data is required. The main objective of this paper is to describe a SAS macro that can be used to estimate separate covariate effects on multiple, correlated binary outcomes. We demonstrate the utility of the macro by applying it to fit a trivariate logistic regression model using GEE where the three correlated longitudinal outcomes of interest include whether a subject had a problem-oriented visit, a dental cleaning, or a routine check-up, or some combination thereof. All three outcomes were measured at four 6-monthly intervals (0-24 months). Estimates from the trivariate logistic regression model are compared to results obtained by fitting three separate binary longitudinal models using GEE for each oral health outcome. The odds of having a problem-oriented visit were greater for males compared to females as estimated from the multivariate model (P = 0.0407), but the odds were not significant in the univariate model (P = 0.0641). The multivariate model also aided in confirming expected results that consistent regular attenders (compared to consistent problem-oriented attenders) had greater odds of having received dental cleaning and check-ups relative to having problem-oriented visits (chi2 = 33.47, P < 0.01), and that those with broken teeth or broken filling (compared to those without) are at greater odds of having a problem-oriented visit relative to having dental cleaning or checkups (chi2 = 34.12, P < 0.01 and chi2 = 17.11, P < 0.01).