The Sustainable Development Goals Report 2026

Released 7 July 2026 · UN DESA · New York
Published
7 July 2026, ahead of the High-Level Political Forum
Publisher
UN Department of Economic and Social Affairs (DESA), with 50+ international and regional organizations
Data cutoff
Latest available data as of June 2026; 139 SDG targets assessed against trend data
This year’s lens
“SDG monitoring in the AI era”: renewing official statistics for people, trust and sovereignty
Launch
Press conference by Deputy Secretary-General Amina J. Mohammed and UN DESA

Every year the UN publishes one number that matters more than any speech: the share of the Sustainable Development Goals actually on track. For 2026, at roughly two-thirds of the way to the 2030 deadline, that number is 36%. This is the report SDGCounting exists to read, so here is what it says, and the quieter story about the data underneath it.

The headline

Of the 139 SDG targets with enough trend data to assess, just 36% are on track or making moderate progress. Nearly half (49%) are advancing too slowly, and 15% have slipped below their 2015 baselines. The report grades every assessable target on a five-point scale:

The Secretary-General’s framing is that the goals are working, but not fast or evenly enough. The gains, he writes, “are real and span the Goals,” while progress “remains uneven and insufficient in the face of serious headwinds”: aid in retreat, record debt, conflict at its highest in decades, and a climate line now crossed.

António Guterres, UN Secretary-General (foreword): “With less than five years to achieve the Sustainable Development Goals, this report offers a clear-eyed measure of our collective progress — and the urgent work ahead.”

Where the Goals stand

The report is a mix of genuine decade-long wins and stalled or reversing fronts. A selective read:

The five goals under in-depth review at this year’s High-Level Political Forum tell the same split story. Water (SDG 6): 74% now have safely managed drinking water, up from 68% in 2015. Energy (SDG 7): electricity reaches 92%, and renewables are 18% of final energy consumption. Industry (SDG 9): 5G covers 55% of the world but only 4% of people in low-income countries. Cities (SDG 11): the slum population has climbed to 1.16 billion, nearly a quarter of all urban residents. Partnerships (SDG 17): official development assistance fell a record 23.1% to $174.3 billion, developing-country external debt hit a record $8.9 trillion, and about 74% of the world is now online, up from 40% in 2015.

The counting story: data in the AI era

This edition carries a feature essay on the state of measurement itself, which is where SDGCounting pays closest attention. The decade’s least-celebrated achievement is the statistical machinery behind the Goals. The SDG database has grown from 115 indicators in 2016 to 233 in 2026, and from roughly 330,000 data records to 3.2 million. A decade ago the world lacked data for half of the indicators; today every indicator has an internationally agreed method, and the share of indicators with good country coverage has risen from about a third to nearly 70%.

The gaps are just as telling. Gender equality, sustainable cities, climate action, and peace and justice each still have fewer than one in three indicators with enough trend data to judge, and many countries cannot disaggregate by income, disability, or migration status, which is exactly what it takes to see who is being left behind. The report is candid that the aid collapse now threatens the measurement itself: of 135 countries with national statistical plans, only 59% are fully funded, and in sub-Saharan Africa just 15%.

AI runs through the essay as both tool and hazard. It can extend statistics to undercounted populations, but when official numbers are missing, large language models “fill gaps through extrapolation, interpolation or synthetic generation, producing figures that may appear authoritative while masking significant uncertainty.” The report’s prescription is not a new rulebook but capacity, financing and independence, with national statistical offices asserting their role as “the verifiable source of record.” Two tracks carry the response: the Sevilla Commitment on financing and the Medellín Framework for data systems.

The launch

Deputy Secretary-General Amina J. Mohammed presented the report with UN DESA, and put the crisis where the data does: not in the goals, but in the means to reach them.

Amina J. Mohammed, UN Deputy Secretary-General: “In the same year that military spending reached $2.9 trillion, development assistance suffered its steepest fall on record, while the annual SDG financing gap still stands at $4 trillion a year.”

She was emphatic that the framework is sound. “This is not a failure of the goals,” she said; where they are “backed by political will and resources, they do deliver.” On measurement, she made the point SDGCounting keeps returning to: “We measure more, we measure better, and with greater precision than we’ve ever done. And what the data shows is that progress has not reached everyone everywhere.” The Goals, she argued, remain achievable “if we choose to act together with greater urgency, scale, solidarity, and resolve, but that choice must be made now.”

Our read: 36% on track sounds like failure and reads like one in a headline, but the more useful number this year is the denominator. Only 139 of the targets can even be assessed, and the goals that measure people most at risk, on gender, cities, climate and peace, are the ones the world can least see. The report’s quiet warning is that the aid cuts are now hitting the statistics themselves, so the danger is not only slower progress but a dimming of the very instrument that tells us where we stand. UN DESA put it plainly: lives improved over the decade, but progress remains insufficient.

Watch & read

Figures are drawn from the report as published; the report notes minor internal rounding differences (for example, 3 million versus 3.2 million data points) and these are reproduced as written. Quotations from the launch are lightly edited from an automated (Otter.ai) transcript of the UN Web TV recording and should be read as close paraphrase.