Small rotary compressors are used in domestic heat pump appliances, for example, in domestic dehumidifiers and heat pump clothes dryers. Compressor performance curves provided by the manufacturer can be based on testing at relatively high ambient temperatures, in some cases as high as 35°C. This can be much higher compared with the ambient temperature in which the compressor operates when, for example, it is installed in a domestic dehumidifier which can operate in ambient temperatures as low as 10°C. We have developed a compressor calorimeter to test a small R134a rotary compressor extracted from a commercial domestic dehumidifier and use this to measure compressor performance parameters including the isentropic and volumetric efficiencies and the compressor heat loss fraction. The performance testing has been carried out at ambient temperatures 10°C, 15°C, 20°C and 25°C for a fixed relative humidity of 70% to compare how the compressor performance varies with the ambient temperature, and to determine how well the compressor performs outside of the performance envelope provided by the manufacturer. The results show that isentropic and volumetric efficiency of these small compressors is relatively insensitive to variation in ambient temperature, even outside of the performance envelope provided by the manufacturer. However, the compressor heat loss fraction can, on average, double from 15% to 30%, between operation at ambient 25°C and ambient 10°C. The data obtained in this work is used to construct compressor sub-models for certain ambient temperatures. We show how these sub-models can be used to improve a domestic dehumidifier model for operation at low ambient conditions within the evaporator frosting regime and good agreement is obtained between experimental and simulated data. The authors are not aware of a domestic dehumidifier model designed to work at ambient temperatures within the frosting regime.