Objectives: We aimed to investigate levels of the macrophage-specific marker, sCD163, in cerebrospinal fluid and plasma in patients with Lyme neuroborreliosis. We tested the diagnostic value of CSF-sCD163 and ReaScan-CXCL13 and analyzed if plasma-sCD163 could monitor treatment response. Methods: An observational cohort study: Cohort 1—Cerebrospinal fluid from adults with neuroborreliosis (n = 42), bacterial meningitis (n = 16), enteroviral meningitis (n = 29), and controls (n = 33); Cohort 2—Plasma from 23 adults with neuroborreliosis collected at diagnosis, three, and six months. sCD163 was determined using an in-house sandwich ELISA. ReaScan-CXCL13 measured semiquantitative concentrations of CXCL13, cut-off ≥ 250 pg/ml diagnosed neuroborreliosis. Receiver Operating Characteristics analyzed the diagnostic strength. A linear mixed model including follow-up as categorical fixed effect analyzed differences in plasma-sCD163. Results: CSF-sCD163 was higher in neuroborreliosis (643 µg/l) than in enteroviral meningitis (106 µg/l, p < 0.0001) and controls (87 µg/l, p < 0.0001), but not bacterial meningitis (669 µg/l, p = 0.9). The optimal cut-off was 210 µg/l, area under the curve (AUC) 0.85. ReaScan-CXCL13 had an AUC of 0.83. Combining ReaScan-CXCL13 with CSF-sCD163 increased AUC significantly to 0.89. Plasma-sCD163 showed little variation and was not elevated during the 6 months of follow-up. Conclusion: CSF-sCD163 is diagnostic for neuroborreliosis with an optimal cut-off of 210 µg/l. Combining ReaScan-CXCL13 with CSF-sCD163 increases AUC. Plasma-sCD163 cannot monitor treatment response.