Background: Since sensitivity and specificity vary widely between tests, SARS-CoV-2 serology results should be interpreted with caution. Methods: The study included serum samples from patients who had recovered from COVID-19 (n = 71), individuals vaccinated against SARS-CoV-2 (n = 84), and asymptomatic individuals (n = 33). All samples were tested for the presence of binding antibodies (enzyme immunoassay; EIA), neutralizing (NT) antibodies (virus neutralization test; VNT), and surrogate NT (sNT) antibodies (surrogate virus neutralization test; sVNT) of SARS-CoV-2. Results: SARS-CoV-2-binding antibodies were detected in 71 (100%) COVID-19 patients, 77 (91.6%) vaccinated individuals, and 4 (12.1%) control subjects. Among EIA-positive samples, VNT was positive (titer ≥ 8) in 100% of COVID-19 patients and 63 (75.0%) of the vaccinated individuals, while sVNT was positive (>30% inhibition) in 62 (87.3%) patients and 59 (70.2%) vaccinated individuals. The analysis of antibody levels showed a significant moderate positive correlation between EIA and VNT, a moderate positive correlation between EIA and sVNT, and a strong positive correlation between VNT and sVNT. The proportion of positive sVNT detection rate was associated with VNT titer. The lowest positivity (72.4%/70.8%) was detected in samples with low NT titers (8/16) and increased progressively from 88.2% in samples with titer 32 to 100% in samples with titer 256. Conclusions: sVNT appeared to be a reliable method for the assessment COVID-19 serology in patients with high antibody levels, while false-negative results were frequently observed in patients with low NT titers.