Abstract Cryogenic turbo-expander is most significant equipment to provide cooling energy in system and its working condition is always deviated from design point. A mathematic prediction method study is carried out to estimate turbo-expander off-design performance. Computational iterative loop is compiled by Matlab, dimensionless mass flow rate equation of mean streamline and novel loss correlation are applied to quantitatively describe the flow expansion through turbine ducts. Cryogenic turbo-expander performance is evaluated by total-to-static efficiency. According to velocity ratio, the effect of pressure ratio, inlet temperature, rotational speed variation to turbine performance is analyzed, and the predicted performance map is plotted. Meanwhile, an experimental study is conducted under off-design condition. Temperature, pressure, rotation speed and volume flow rate are collected, total-to-static efficiency are calculated from turbine inlet and outlet states over pressure ratio range of 2.4–3.4, and the tested rotation speed range is set from 52,000 to 60,000 rpm. Turbine efficiency in different pressure ratio range are categorized and plotted with velocity ratio to validate against computational predicted characteristic. With experimental comparison, this off-design performance mathematic code can predict turbine real operation well.