Abstract Study question Is it useful to avoid sperm centrifugation in laboratory routine work to improve sperm quality and reproductive outcome in Assisted Reproduction Techniques (ART)? Summary answer Exclusion of sperm centrifugation for sperm selection using neat sperm samples (IO-lix), increases sperm quality in the collected subpopulation decreasing miscarriage rate after using ICSI. What is known already Inclusion of sperm centrifugation in ART is an aggressive intervention for sperm selection with ineludible production iatrogenic damage affecting sperm integrity. The application of IMSI, PICSI or microfluidic devices avoid sperm centrifugation and may improve the quality of the subsample obtained. However, these methodologies may result time consuming, expensive or producing poor results when the quality of the sperm is limited. We have already shown that a modified swim-up avoiding centrifugation (called IO-lix) is a low-cost and efficient alternative to microfluidic devices, recovers 100 times more concentration and reduces sperm DNA fragmentation with no significant differences to other methodologies. Study design, size, duration This is a retrospective study from 2018 to 2020 which includes patients with an average of age of 38.2 years using their own oocytes with ICSI as fertilization technique. Two aleatory groups of patients were made: Group 1: 88 cycles with 503 fertilized oocytes and 206 blastocysts were obtained with sperm samples processed by IO-lix and Group 2: 303 cycles, 1451 fertilized oocytes and 591 blastocysts using a standard “swim up” technique to process sperm. Participants/materials, setting, methods A total of 391 ICSI cycles were included in this retrospective study. The male factor was similar in both groups and they showed altered SDF previously to the cycle. We compared data of the motility and SDF of sperm samples before and after applying IO-lix and we analyzed by X2 contingence test differences on miscarriage rates between groups 1 and 2. Main results and the role of chance General sperm parameter changes after IO-lix showed that averaged sperm concentration observed in neat ejaculated samples was 62M/SD=46.4. Values obtained after IO-lix in the same samples were 12.3M/SD8.0. Averaged sperm motility in neat samples was 54%/SD=9.3 and 70.9%/SD=13.2 after IO-lix. Finally, sperm DNA fragmentation in neat samples was 35.8%/SD17.3, while these values decreased to 9.2%/SD=3.9 after IO-lix. About reproductive outcome results, significant differences were not obtained on the development to blastocyst stage rate comparing both groups (X2=0.003; p value = 0.954; Alpha 0.05). In the case of IO-lix processed samples, the pregnancy rate was 59.42% in Group 1 and 44.72% in Group 2 (X2=0.651; p value =0.419; Alpha 0.05). A total of 9 miscarriages of 41 clinical pregnancies (21.95%) were observed after IO-lix, while this number increases to 59 out of 123 clinical pregnancies, which means the 47.96% of the embryo transfers, when “swim-up” was used. In this case significant differences were obtained (X2=3.935; p value = 0.0.047; Alpha 0.05). Limitations, reasons for caution Being a pilot study aimed to understand the results of IO-lix in ART, correlations have not been stablished between the levels of sperm improvement after IO-lix and paired results of ART. This study would be necessary, specially to identify the possible origin of miscarriage associated to the male factor. Wider implications of the findings: Elimination of sperm centrifugation using a combined strategy of gradients and “swim-up” for sperm isolation, reduce miscarriage rate and produce equivalent results of blastocyst development to those obtained with “swim-up”. Being a cost-effective and improving laboratory workload, its use for sperm selection is recommended. Trial registration number Not applicable