There are concerns about central nervous system (CNS)-replication of HIV-1 in patients on boosted protease inhibitors. Purpose of this study was to compare HIV-1 viral loads (VLs) from patients treated with only boosted dual protease inhibitor (bdPI), versus combination antiretroviral therapy (cART group), containing two nucleoside analogue reverse transcriptase inhibitors (NRTI) and a third partner. All patients from a large German HIV-treatment cohort with available medication, clinical and demographic data, including results from simultaneous HIV-1 viral load (VL) assessments in cerebrospinal fluid (CSF) and blood plasma, were retrospectively evaluated as controlled cross-sectional study. CSF had been obtained from patients with variable neurological symptoms during 2005-2014. Statistical analysis comprised nonparametric tests, regression and correlation techniques accounting for undetectable quantifications. Statistical analysis comprised nonparametric tests, regression and correlation techniques accounting for undetectable quantifications. Overall, 155 patients were evaluable (bdPI: 24; cART: 131). At time of CSF-collection, both groups were comparable in age, gender, CD4-cell counts, or primary HIV-transmission risks, though bdPI patients were clinically more advanced. The proportion of patients with undetectable HIV-1 (<50 copies/ml) in CSF was lower for bdPI group (25 vs 49.6 %; p = 0.026), but similar in plasma (46 vs 41 %). Median CSF-VL was higher in bdPI group (600 vs 50 copies/ml; p = 0.027) and similar in plasma. Mean VL CSF/plasma ratio was 342.91 for bdPI- and 54.48 for cART patients (p < 0.001). Pearson's regression analysis revealed a trend for an elevated VL-ratio over time within bdPI group. HIV-1 replication was higher and more frequently detectable in CSF from bdPI patients, indicating a worse CNS penetration effectiveness of used boosted PI. Within bdPI group, measured CNS-viral replication was increasing over time, suggesting an over time impaired HIV-1 suppression in CSF.