Understanding carbon and oxygen isotope systematics in cave environments is a prerequisite for the interpretation of stable isotopes in speleothem-based paleoclimate records. Here we present a series of experimental data collected under laboratory conditions with controlled temperature, relative humidity, drip water chemistry, flow rate and cave pCO2, simulating the growth of speleothems in natural cave settings. Drip water with high pCO2 and low calcium concentration (Ca2+ = 2 mmol/L) flowed along a three-step glass path, similar to a stalactite-stalagmite-pool route in natural caves, forming a thin water film that allowed CO2 degassing and CaCO3 precipitation as a result of the pCO2 gradient between the drip water and ambient cave atmosphere. The experiments were conducted at 15, 25 and 32 °C and flow rates of 700, 270 and 125 ml/d.The growth rate of calcite on the stalagmite-like settings increases linearly with increasing flow rate and/or temperature. The δ13CCc and δ18OCc of calcite formed on the stalagmite-like settings increases with decreasing flow rate (corresponding to increasing exposure time of water) at a given temperature, indicating non-equilibrium isotope effects between calcite, water and dissolved inorganic carbon (DIC). Nevertheless, these non-equilibrium isotope effects still display regular temperature dependence under a constant flow rate. This suggests that non-equilibrium isotope effects in natural stalagmites might be used to provide useful qualitative paleoclimate information (such as differentiating wet/dry and warm/cold climate conditions). The non-equilibrium carbon and oxygen isotope effects in the stalagmite-like settings were most likely caused by rapid CO2 degassing and CaCO3 precipitation that rapidly consume the available DIC pool in the thin water film. Furthermore, CO2 exchange between DIC and cave atmosphere quickly amplified the observed non-equilibrium carbon isotope effects in the precipitated calcite.In the pool-like settings, calcite was buffered by oxygen isotope exchange between DIC species and water, and slowly precipitated at or near to oxygen isotopic equilibrium with the temperature dependence of 1000ln18αCc-H2O = 18.33 (103/T) – 33.31 regardless of flow rate. This fractionation relation agrees with that determined by Kim and O’Neil (1997) when a newly recommended value for the acid fractionation factor for calcite is used (i.e., Kim et al., 2015). Carbon isotope fractionation between calcite and bicarbonate was temperature-independent between 15 and 32 °C and the average magnitude was 1000ln13αCc–HCO3− = 1.7 ± 0.7‰. Observed variability of 1000ln18αCc-H2O in modern calcite speleothems from natural cave settings lies in the range predicted by this study, between the predicted maximum non-equilibrium deviation at the stalagmite-like settings and the equilibrium 1000ln18αCc-H2O achieved in the pool-like settings.