The recent development of miniature particle detectors stems from the demanding need to monitor/measure particles, especially nanoparticles, at the personal level for epidemiological studies or studies investigating the interaction among genes and environmental factors, including particulates. Light scattering and electrical mobility techniques have been implemented in these mini-devices for monitoring submicron particles. The presence of large particles in the sampling stream, however, affects the performance of these mini-detectors. Prototype mini-cyclones were thus developed as a size selective inlet for mini-particle detectors. In this study two “quarter-sized” mini-cyclones were designed to remove particles larger than 1.0 and 0.3μm at a flow rate of 0.3lpm, and their performance was experimentally evaluated. The performance of prototypes was also compared with that of existing personal sampling cyclones. Further, empirical models to estimate the performance of the prototype mini-cyclones (i.e., the 50% cutoff particle size and pressure drop) were also established. The developed linear regression model can thus serve as the tool for the future design of mini-cyclones with the similar size and configuration.