Dot-blot immunoassays are widely used for the user-friendly detection of clinical biomarkers. However, the majority of dot-blot assays have only limited sensitivity and are only used for qualitative or semiquantitative analysis. To overcome this limitation, we have employed labels based on photon-upconversion nanoparticles (UCNPs) that exhibit anti-Stokes luminescence and can be detected without optical background interference. First, the dot-blot immunoassay on a nitrocellulose membrane was optimized for the quantitative analysis of human serum albumin (HSA), resulting in a limit of detection (LOD) of 0.19 ng/mL and a signal-to-background ratio (S/B) of 722. Commercial quantum dots were used as a reference label, reaching the LOD of 4.32 ng/mL and the S/B of 3, clearly indicating the advantages of UCNPs. In addition, the potential of UCNP-based dot-blot for real sample analysis was confirmed by analyzing spiked urine samples, reaching the LOD of 0.24 ng/mL and recovery rates from 79 to 123%. Furthermore, we demonstrated the versatility and robustness of the assay by adapting it to the detection of two other clinically relevant biomarkers, prostate-specific antigen (PSA) and cardiac troponin (cTn), reaching the LODs in spiked serum of 9.4 pg/mL and 0.62 ng/mL for PSA and cTn, respectively. Finally, clinical samples of patients examined for prostate cancer were analyzed, achieving a strong correlation with the reference electrochemiluminescence immunoassay (recovery rates from 89 to 117%). The achieved results demonstrate that UCNPs are highly sensitive labels that enable the development of dot-blot immunoassays for quantitative analysis of low-abundance biomarkers.