Our objective was to study the effect of lopinavir/ritonavir on cerebrospinal fluid (CSF) viral load as part of an antiretroviral combination treatment for HIV-1 infected individuals, and to determine the steady-state concentrations of lopinavir in CSF in relationship to plasma concentrations. Paired CSF and plasma samples from 12 antiretroviral-naïve HIV-1 infected patients starting combination therapy containing lopinavir/ritonavir were collected at baseline, and during treatment at a first follow-up at median 3.0 months (range 2.6-6.0 months), and at a second follow-up at median 12.1 months (range 6.0-16.5 months). Levels of HIV-1 RNA, CD4+ T-cell count, beta2-microglobulin, neopterin, and lopinavir concentration were analysed. In addition, CSF and plasma lopinavir concentrations in 4 patients already on combination therapy including lopinavir/ritonavir were analysed. Nine of 11 patients had undetectable viral load in CSF and 5/11 in plasma at the first follow-up. At the second follow-up 7/7 had undetectable viral load in CSF and 9/9 in plasma. Intrathecal immunoactivation, measured by neopterin and beta2-microglobulin, decreased significantly both in CSF and serum. The total CSF concentrations of lopinavir were of the same order of magnitude as the unbound concentrations in plasma. Lopinavir mean (+/-SD) concentrations were 42.1 (+/-31.5) nM in CSF and 52.7 (+/-25.2) nM unbound in plasma. We found that antiretroviral combination therapy including lopinavir/ritonavir substantially decreases the viral load, both in CSF and plasma, as well as the intrathecal immunoactivation, measured by beta2-microglobulin and neopterin. CSF concentrations of lopinavir were low, but probably sufficient to have a virological effect.