an organizationInternational Committee for Monitoring Assisted Reproductive Technologies

The world registry of assisted reproduction: an independent non-profit that describes itself as having "taken the leading role in the development, collection and dissemination of worldwide data on assisted reproductive technology". It began international data collection in 1989, incorporated in California as a 501(c)(3) in 2003, and is a non-State actor in official relations with the World Health Organization. With WHO it wrote the ICMART ART Glossary, the agreed definitions the field argues inside.

Follow the data upwards and the limits explain themselves. A clinic reports patient-level data to its national or regional registry. ICMART's regional representatives then collect from those registries "non-identifiable tabulated summary data". ICMART never sees a clinic, never sees a patient, and receives whatever the registry chose to send, in the shape the registry already held it.

The scale is real. For cycles conducted in 2017 and 2018, 83 countries provided data — 80 for 2017, 81 for 2018 — from 3,313 reporting clinics in 2017 and 3,254 in 2018. Those returns held 2,913,498 cycles and 671,012 babies for 2017, and 3,303,505 cycles and 728,383 babies for 2018. Adjusting for centres that did not report, ICMART estimates 3,107,188 cycles and 814,588 babies for 2017 and 3,568,635 and 870,814 for 2018. The report names its own limits: "Data and analysis presented are dependent on the quality and completeness of the data submitted to ICMART, which varies globally", with thin coverage from the Middle East and incomplete coverage from parts of Asia, and a change in how delivery rates are calculated that blocks comparison with earlier ICMART reports.

Inference — three problems stack, and they stack against exactly the reader this directory has. A country appears only if it has a registry and that registry chooses to report, so the destinations with the least regulation are the ones least likely to be in the table. What a registry reports is what it already collects, on its own definitions of a cycle, a transfer and an outcome — which is why the glossary exists, and the glossary narrows the differences rather than abolishing them. And the unit is the country: a patient choosing between two clinics abroad can read a national aggregate several years old, and nothing about either clinic. The world total is the best figure the field has, and it still cannot answer the question that makes someone buy a plane ticket.

Standard reporting forms go to national and regional registries; ICMART's regional representatives gather the returns; analysis has been done by the National Perinatal Epidemiology and Statistics Unit, with built-in and manual validation checks and missing data imputed using weighted average cancellation rates. Preliminary world reports are presented at ESHRE's annual meeting; final world reports are published in Human Reproduction, several years behind the cycles they describe.

As of 17 September 2026 the most recent full world report covers cycles conducted in 2017 and 2018 and was published in Human Reproduction in 2025; a preliminary report for 2021 was presented at the ESHRE meeting in Paris in 2025. The lag between a cycle and its appearance in a world figure runs to several years, which is longer than most patients wait to travel.

Reporting to ICMART is voluntary. A country absent from a world report has not been shown to perform no cycles; it has been shown not to have reported any.