Objective. To determine the baseline factors predictive of poor compliance with study visits in a longitudinal multiethnic lupus cohort study. Methods. Patients with systemic lupus erythematosus (n = 344) representing a total of 2,069 potential study visits were studied. Of the participants, 24.4% were Hispanic, 43.3% African American, and 32.3% Caucasian. Noncompliance was defined as missing 2 or more study visits. For the purpose of these analyses, visits completed only by review of medical records were considered missing. Baseline socioeconomic-demographic, clinical, and psychosocial features between compliant and noncompliant patients were compared. Variables with P < 0.10 were then entered into a multivariable logistic regression analysis with compliance being the dependent variable. Results. There were 178 compliant and 166 noncompliant patients. Noncompliant patients were more likely to be young, unmarried, of African American ethnicity, live closer to the medical centers, and have longer disease duration and greater disease activity as assessed by the physician than the compliant patients. In the multivariable model, longer disease duration (P = 0.010), higher level of disease activity (P = 0.009), and shorter distance to travel to study visits (P = 0.046) were predictors of noncompliance; their odds ratios and confidence intervals were below 1. Conclusions. We have identified baseline patient characteristics that may predict noncompliance with study visits (disease duration, disease activity, and distance to the medical center). This information will serve as the basis for developing interventions to curtail noncompliance. Our data may have applicability in other lupus cohort studies.