In vitro semen purification techniques have been developed that seek to mimic the in vivo selection process, in order to generate the highest possible chance of oocyte fertilization following artificial insemination (Morrell and Rodriguez-Martinez. Vet Med Int. 2011:894767). Since this practice began, there have been many methods developed in an attempt to isolate functional spermatozoa for AI, yet only one method is routinely used – density gradient centrifugation using commercial preparations such as EquiPure. Here, we have developed a novel method of isolating functional spermatozoa and compared sperm quality to spermatozoa isolated with EquiPure. The AI-Port system (Memphasys, Sydney, Australia), comprises a disposable cartridge consisting of a sample chamber – where extended semen is deposited, and a harvest chamber – where isolated spermatozoa is extracted. These chambers are separated by a 5 µm nylon woven membrane, through which highly motile spermatozoa are able to swim from the sample chamber into the harvest chamber over 20 min, leaving behind seminal plasma and other cell types such as leukocytes. Spermatozoa were collected from miniature pony stallions, and immediately extended with EquiPlus semen extender. The ejaculate was split, and spermatozoa were isolated into BWW media with EquiPure (as per manufacturer's instructions; Nidacon, International), or the AI-Port system. Comparisons were conducted between spermatozoa isolated with the AI-Port and EquiPure, which revealed that the yield of AI-Port samples was significantly lower (25.5±5.9%) incomparison to EquiPure samples (44.9±10.2%; Kruskal-Wallis), potentially solvable with prototype improvement. Viability (Eosin-nigrosin) of AI-Port samples was significantly higher than EquiPure samples (94.4±1.1 vs 81.5±2.1%; P≤0.01; ANOVA; respectively). In regard to motility, total motility for AI-Port samples was significantly higher than EquiPure samples (93±1.3 vs 89.5±2.9%; P≤0.05; ANOVA; respectively). Whilst there was no significant difference for progressive motility between AI-Port and EquiPure samples (45.7±3.2 vs 45.6±3.4%; ANOVA; respectively). Both samples were higher than the progressive motility of the initial non-purified ejaculate (30±4.1%; P≤0.05). Mitochondrial superoxide was measured with flow cytometry and MitoSOX red. Likewise, there was no significant difference in superoxide levels between AI-Port and EquiPure samples (10.2±3.1 vs 5.6±0.8%; ANOVA; respectively). However, both samples had a reduced level of superoxide generation compared to the initial non-purified ejaculate (32.5±20.2%; P≤0.05). The DNA integrity was measured with the Halo test, which confirmed AI-Port samples had less DNA damage than EquiPure™ samples (2.0±0.5 vs 10.2±2.1%; ANOVA; P≤0.01; respectively). In summary, the AI-Port effectively isolates high-quality spermatozoa at a comparable or superior level to EquiPure. Its cost-effectiveness and user-friendliness make it a viable option for improving artificial insemination success rates in breeding programs.