HIV-1 viral load (VL) testing is a standard component of HIV care. We examined the use and predictors of VL testing in drug users, a group at risk for problematic care. Using 1996 to 1998 New York State (NYS) Medicaid files, we studied drug users who had been enrolled >10 months, had been prescribed antiretroviral agents in 1997 and 1998, and who had undergone any VL testing in 1997. Our outcome was regular VL testing shown by two or more paid claims for this test in 1998. Patterns of care in 1997 were defined as: regular source of medical care (>35% of visits to one provider), and/or regular drug treatment of >6 months, or neither. We counted visits in 1997 to a provider offering HIV-focused care. Adjusted odds ratios (AORs) of VL testing were assessed. Of 3131 drug users, 73.9% had at least one VL test, whereas 56.2% had two or more VL tests in 1998. The AORs of two or more VL tests were increased for those with regular drug abuse care alone (AOR, 1.50; 95% confidence interval [CI], 1.21–1.84) or with regular medical care (AOR, 1.27; 95% CI, 1.03–1.57) versus those with neither. HIV-focused care was positively associated with two or more VL tests (AOR, 1.38; 95% CI, 1.05–1.81 for 1–3 visits; AOR, 1.94; 95% CI, 1.50–2.51 for four or more visits). We found that nearly half this cohort of drug users did not have regular VL testing. Drug users with HIV-focused care or with regular drug treatment are more likely to have regular VL testing.