Abstract Despite being identified nearly a century ago, the diffusion absorption refrigeration (DAR) cycle has received relatively little attention. One of the strongest attractions of the DAR cycle lies in the fact that it is thermally driven and does not require high value work. This makes it a prime candidate for harnessing low-grade heat from solar collectors, or the waste heat from stationary generators, to produce cooling. However, to realize the benefits of the DAR cycle, there is a need to develop an improved understanding of how design parameters influence its performance. In this vein, this work developed a new, validated, parametric model that can be used by designers and practitioners to examine the performance of the DAR cycle for a range of operating conditions. The results showed that the cycle’s performance was particularly sensitive to several factors: the rate of heat added and the temperature of the generator, the effectiveness of the gas and solution heat exchangers, the mass flowrate of the refrigerant and the type of the working fluid. It was shown that it can deliver good performance at low generator temperatures if the refrigerant mass fraction in the strong solution is made as high as possible. Moreover, it was shown that a H2O–LiBr working pair could be useful for achieving cooling at low generator temperatures.