Experimentally metasomatised monazite was studied in terms of preservation of U-Pb and Th-Pb ages during alkali-bearing fluid-induced alteration over a broad range of temperature conditions 250–750 °C. Starting materials for experiments included Burnet monazite (Concordia age 1100.5 ± 11.6 Ma, 2σ), albite, K-feldspar, biotite, muscovite, SiO2, CaF2, Na2Si2O5 and H2O. Monazite from experiments at 250–550 °C is partially replaced by secondary REE-rich fluorapatite [(Ca,LREE,Si,Na)5(PO4)3F], fluorcalciobritholite [(Ca,REE)5(SiO4,PO4)3F] and REE-rich steacyite [K1−x(Na,Ca)2(Th,U)Si8O20], and developed patchy zoning, whereas partial replacement by fluorcalciobritholite and cheralite [CaTh(PO4)2] occurred at 650 and 750 °C, with no signs of compositional alteration based on EPMA data and BSE imaging. Raman microspectroscopy results show narrowing of the ν1(PO4) stretching band in unaltered domains, which indicates advancing annealing of the monazite structure with increasing temperature, and narrow ν1(PO4) band with low FWHM values in altered domains. TEM investigations revealed that unaltered domains of monazite from experiments at 250–550 °C have mottled diffraction contrast, similar to the starting Burnet monazite, which indicates low to moderate degree of metamictization. On the contrary, the altered domains of monazite (patchy zones) show no mottled contrast, suggesting an ordered crystalline structure. TEM imaging demonstrated low degree of metamictization in monazite from the experiment at 650 °C; fluid-aided alteration along the cleavage planes resulted in the development of nanoporosity or partial replacement by fluorcalciobritholite and cheralite. Monazite from the experiment at 750 °C has crystalline structure with no signs of metamictization and shows significant development of nanoporosity and formation of secondary cheralite nanocrystals across the grain. For comparison, TEM and Raman evaluation of xenotime from similar experiments at 350 and 650 °C revealed that both starting xenotime and xenotime from experimental products are crystalline with no signs of radiation damage or fluid-induced alteration affecting internal domains on submicron scale, which could result in compositional alteration of the xenotime.The unaltered domains of monazite from runs at 250–550 °C yielded U-Pb and Th-Pb dates similar to the age of Burnet monazite, whereas altered domains yielded discordant dates due to various degree of Pb-loss (up to 99.4%). Linear regressions on the Concordia diagrams show lower intercept ages from −266 ± 160 Ma (run 350 °C, 200 MPa) to −1 ± 48 Ma (450 °C, 800 MPa), which reflect the “true age” of experimental alteration. The monazite from runs at 650 and 750 °C yielded data indicating initial disturbance of the U-Th-Pb system, ranging from 8.4% Pb-gain to 18.6% Pb-loss. Linear regressions with lower intercepts of −53 ± 420 Ma and −55 ± 610 Ma roughly correspond to the timing of the experiments. Furthermore, LA-ICPMS results demonstrate discrepancy between Th-Pb and U-Pb dates suggesting higher mobility of 208Pb than that of 207Pb and 206Pb.To summarize, TEM and Raman data indicate increasing annealing of the radiation damaged monazite with increasing temperature. Alteration processes induced by alkali-bearing fluid can result in recrystallization of monazite and various degrees of the age disturbance at temperatures 250–550 °C, whereas isotopic U-Th-Pb microanalysis provide an opportunity to constrain the age of the metasomatic processes as the lower intercept in the Concordia diagram. The particular importance of this study lies in submicron alteration of monazite at 650–750 °C induced by alkali-bearing fluid and/or melt, which remains unnoticed using common electron microscopy BSE imaging. Such alteration, however, induces substantial disturbance of U-Pb and Th-Pb ages, which can cause misinterpretations in reconstructions of geological processes.