Point-of-service (POS) HIV testing in sexually transmitted infection (STI) clinics is one public health strategy to increase knowledge of serostatus and to facilitate entry into care. Variation has been reported in clients' views of test reliability and rates of test acceptance. Our objective was to characterize STI clinic patients' choice of POS versus conventional testing (enzyme-linked immunosorbant assay [ELISA] followed by Western blot, with results in 1 week) in Baltimore, Maryland (a high-prevalence city) when both were offered (May through August 2008), then to compare rates of engaging in care. Odds ratios (OR) with 95% confidence intervals (CI) described factors associated with test type choice, as well as HIV test type with entrance into care. The overall prevalence of HIV among testers was 1.1% (60/5101). Those reporting receptive anal sex (OR 1.4; 95% CI 1.1-1.7), illicit drug use (OR 1.3; 95% CI 1.0-1.6), or an HIV-positive sexual contact (OR 1.5; 95% CI 1.0-2.2) were more likely to choose POS testing, as were those who had been tested for HIV previously (OR 1.3; 95% CI 1.1-1.5). Hispanics were less likely to choose POS testing (OR 0.6; 95% CI 0.4- 0.7). Entry into care was low in both categories of test takers (52% in POS testers versus 42% in conventional testers, p = 0.58). Patients at the highest risk for HIV preferred POS testing in STI clinics. Strengthening linkage to care is important for optimizing outcomes of HIV-positive patients presenting to STI clinics.