Gmail alone blocks nearly 15 billion unwanted emails every day, according to Google's own October 2023 announcement. That works out to about 174,000 spam, phishing and malware messages stopped every second, by one provider. Across all providers, 44.99% of the world's email was spam in 2025, according to Kaspersky's annual spam and phishing report, which means roughly four spam messages travel for every five legitimate ones.
So how many spam emails are sent per day in total? Nobody counts that directly. Every published figure is one vendor's telemetry, or one vendor's spam share multiplied by someone's forecast of total email, and the answers people quote swing by tens of billions depending on whose numbers they multiply. A single "160 billion a day" headline hides more than it tells.
This page shows the range instead: what each primary source measured, how, and in which year. It covers volume, the share of email that's spam, where spam comes from, what it carries and what it costs. Every number links to its original publisher. Cite freely with a link.
Spam email statistics at a glance
| Category | Statistic | Source |
|---|---|---|
| Share | 44.99% of global email traffic was spam in 2025, down 2.28 points from the year before | Kaspersky, 2026 |
| Share | Spam peaked at 52.87% of all email in November 2025 | Kaspersky, 2026 |
| Share | 36.9% of email received by businesses was unwanted: 20.5 billion of 55.6 billion messages | Hornetsecurity, 2024 |
| Share | 5.6% of emails Cloudflare analysed in 2025 were malicious | Cloudflare, 2025 |
| Volume | Gmail blocks nearly 15 billion unwanted emails a day and stops more than 99.9% of spam | Google, 2023 |
| Volume | Microsoft screens 5 billion emails a day for malware and phishing | Microsoft, 2025 |
| Origin | Russia sent 32.50% of spam, China 19.10%, the United States 10.57% | Kaspersky, 2026 |
| Origin | Botnet command-and-control activity rose 24% in the second half of 2025 | Spamhaus, 2026 |
| Payload | Kaspersky blocked 144,722,674 malicious email attachments in 2025 | Kaspersky, 2026 |
| Payload | Malware-laden email rose 131% year over year | Hornetsecurity, 2025 |
| Payload | 1,069,681 unique phishing sites were reported in Q2 2026, up 10.1% | APWG, 2026 |
| Cost | Business email compromise cost US victims $3.05 billion in 2025 | FBI IC3, 2026 |
| Cost | Phishing and spoofing was the most-reported internet crime: 191,561 complaints | FBI IC3, 2026 |
| Breaches | Phishing was the initial access vector in 16% of breaches | Verizon DBIR, 2026 |
How many spam emails are sent per day?
No organisation publishes a verified worldwide count of spam emails sent per day; the largest single-provider figure is Gmail's nearly 15 billion unwanted emails blocked daily (Google, 2023). Everything else is either a provider's own volume or a share that has to be multiplied by an estimate of total email.
| Statistic | Source |
|---|---|
| Gmail's AI defences block nearly 15 billion unwanted emails every day and stop more than 99.9% of spam, phishing and malware from reaching inboxes. Google published this in October 2023, so the number is dated. | Google, October 2023 |
| Microsoft screens 5 billion emails a day for malware and phishing, alongside more than 100 trillion security signals. | Microsoft Digital Defense Report, October 2025 |
| Proofpoint's threat research is built on more than 3.5 billion emails analysed every day. | Proofpoint Human Factor, August 2025 |
| Hornetsecurity processed more than 72 billion emails between 15 October 2024 and 15 October 2025, over 6 billion a month. | Hornetsecurity, November 2025 |
| In the previous 12-month period, 20.5 billion of the 55.6 billion emails Hornetsecurity reviewed were unwanted, about 56 million a day across its customers. | Hornetsecurity, December 2024 |
| Mimecast flagged more than 9.13 billion threats from 24 trillion data points across 43,000 customers between January and September 2025. | Mimecast, October 2025 |
The daily totals you'll find elsewhere, often between 100 and 200 billion, come from multiplying a spam share like Kaspersky's 44.99% by a forecast of all email sent worldwide. The forecast most of them use comes from Radicati Group, whose site was unreachable when we checked in November 2026, so we don't reprint a total we couldn't verify at the source. The honest answer: well over 15 billion a day, since that's one mailbox provider's blocked mail alone, and probably a large multiple of that.
We'd treat any single "spam emails per day" number as an order of magnitude, not a measurement. Each vendor counts what reaches its own filters, and a blocked message, a rejected connection and a delivered newsletter someone marked as spam are three different things. The 15 billion figure is the one we trust most, because it's a floor from the company that runs the world's biggest inbox.
The share of email that counts as spam is where sources diverge most, which the next section lays out.
What percentage of emails are spam?
Kaspersky measured 44.99% of global email traffic as spam in 2025, the most widely cited share, while business-focused filters report lower figures: Hornetsecurity counted 36.9% of business email as unwanted.
| Statistic | Source |
|---|---|
| 44.99% of all emails sent worldwide in 2025 were spam, a fall of 2.28 percentage points from the previous year (47.27%). | Kaspersky, February 2026 |
| In the Russian web segment, the spam share fell 5.3 points to 43.27%. | Kaspersky, February 2026 |
| 36.9% of all email received by businesses between November 2023 and October 2024 was unwanted. Of that unwanted mail, 2.3% (427.8 million messages) carried malicious content. | Hornetsecurity, December 2024 |
| An average of 5.6% of emails Cloudflare analysed in 2025 were malicious, generally between 4% and 6% for most of the year. This counts malicious mail only, not ordinary spam. | Cloudflare Radar, December 2025 |
| 99.8% of email from the .christmas top-level domain and 99.6% from .lol was spam or malicious. The .best domain had the highest pure spam share at 69%. | Cloudflare Radar, December 2025 |
| Gmail requires every sender to keep the spam rate reported in Postmaster Tools below 0.3%, and recommends staying under 0.10%. | Google sender guidelines, 2026 |
The spread between 5.6% and 44.99% isn't a contradiction. Kaspersky counts all spam across its telemetry; Hornetsecurity counts unwanted mail that reached business gateways; Cloudflare counts only the malicious subset. Each answers a different question about the same flood.
We think the 0.3% Gmail ceiling is the most underrated number on this page. It shows the gap between what spammers send (close to half of all mail) and what a legitimate sender may get flagged for (under one in 300). Real senders get punished for a rate spammers would consider a rounding error.
Spam also isn't spread evenly across the year, which matters if you're reading any single month's figure.
Spam peaks in the fourth quarter
Kaspersky's 2025 data shows spam share rising from about 43.50% for most of the year to 49.26% in the fourth quarter, with November the busiest month at 52.87%.
| Statistic | Source |
|---|---|
| Q4 2025 was the busiest quarter: 49.26% of emails were spam. | Kaspersky, February 2026 |
| Spam peaked at 52.87% in November 2025 and 51.80% in December. | Kaspersky, February 2026 |
| Outside that peak, the spam share held at about 43.50%. | Kaspersky, February 2026 |
| The share of emails Cloudflare processed that contained deceptive links reached as high as 70% in late April and again in mid-November 2025. | Cloudflare Radar, December 2025 |
November's 52.87% is 9.37 points above the rest-of-year average, so more than half of all mail in the shopping season is junk. A sample taken in November will overstate the annual rate; one taken in summer will understate it. That seasonal swing also shapes where spam is sent from.
Where does spam come from?
Russia sent 32.50% of the world's spam in 2025, followed by China at 19.10% and the United States at 10.57%, according to Kaspersky's analysis of outgoing spam by country.
| Statistic | Source |
|---|---|
| Russia's share of spam sent fell from 36.18% to 32.50% in 2025. | Kaspersky, February 2026 |
| China (19.10%) and the United States (10.57%) each gained about 2 percentage points. | Kaspersky, February 2026 |
| Germany rose to fourth place with 3.46%, ahead of Kazakhstan (2.89%) and Hong Kong (2.11%). | Kaspersky, February 2026 |
| Botnet command-and-control activity tracked by Spamhaus rose 24% in July to December 2025. | Spamhaus, January 2026 |
| Remote access trojans made up 42% of the top 20 malware families tied to botnets. | Spamhaus, January 2026 |
| .xyz hosted 21.64% of phishing pages Kaspersky detected in 2025, three times its 2024 share; .com fell from first to fourth (10.52%). | Kaspersky, February 2026 |
Country of origin describes where the sending machine sits, not where the spammer lives. Much spam leaves through hijacked computers and rented servers in botnets, which is why Spamhaus's botnet counts move with spam volume.
From our side of the problem, the source country matters less than the address list. A botnet can send from anywhere, but it can only send to addresses it has. Every address published in plain text on a web page is an entry in someone's list, which is the whole reason a tool like Email Protector exists: it hides the address from scrapers while people still see and click it.
What those messages carry is a bigger worry than where they're sent from.
What do spam emails carry?
Spam increasingly carries links rather than files: Proofpoint found URLs used four times more often than attachments in malicious email in 2025, and Cloudflare found deceptive links in 52% of malicious messages.
| Statistic | Source |
|---|---|
| Kaspersky blocked 144,722,674 malicious email attachments in 2025, about 19 million more than the year before. | Kaspersky, February 2026 |
| Kaspersky's press release puts the rise in malicious and potentially unwanted attachments at 15%; Asia-Pacific accounted for 30% of its email antivirus detections and Europe for 21%. | Kaspersky, February 2026 |
| Kaspersky's anti-phishing system stopped 554,002,207 attempts to follow phishing links in 2025. | Kaspersky, February 2026 |
| Web services were the most impersonated category in phishing (27.42%), ahead of global internet portals (15.89%) and online stores (11.27%). | Kaspersky, February 2026 |
| Deceptive links appeared in 52% of malicious emails in 2025, up from 43%. Brand impersonation rose from 23% to 32%. | Cloudflare Radar, December 2025 |
| URLs are used four times more often than attachments in malicious emails, and Proofpoint saw 4.2 million QR code threats in the first half of 2025. | Proofpoint, August 2025 |
| Malware-laden emails rose 131% year over year, email scams 34.7% and phishing 21%. | Hornetsecurity, November 2025 |
| Phishing made up 77% of all attacks Mimecast saw from January to September 2025, up from 60% in 2024. ClickFix lures grew more than 500% in the first half of the year. | Mimecast, October 2025 |
| 1,069,681 unique phishing sites were reported to the Anti-Phishing Working Group in Q2 2026, up 10.1% from 971,181 in Q1. | APWG, August 2026 |
| Phishing email campaigns reported to APWG rebounded from 35,583 in Q1 2026 to 83,951 in Q2. | APWG, August 2026 |
The shift from attachments to links matters for filters. A file can be scanned once on arrival; a link can point to a clean page at delivery and a credential form an hour later.
Our reading: most spam is still noise, but the noise is the delivery system for the expensive stuff. Hornetsecurity's 2.3% malicious share sounds small until you remember it was 427.8 million messages. A filter that's right 99.9% of the time still lets through a lot when the volume is measured in billions.
Those messages that get through show up in fraud statistics, which the next section covers.
How much do spam and phishing cost?
US victims reported $20.9 billion in internet crime losses in 2025, and business email compromise (BEC) alone accounted for $3.05 billion, according to the FBI's Internet Crime Complaint Center (IC3).
| Statistic | Source |
|---|---|
| IC3 received 1,008,597 complaints reporting $20.877 billion in losses in 2025, a 26% rise in losses from 2024. | FBI IC3, April 2026 |
| Phishing and spoofing was the most-reported crime type with 191,561 complaints, compared with 193,407 in 2024 and 298,878 in 2023. | FBI IC3, April 2026 |
| Reported phishing and spoofing losses rose from $70.0 million in 2024 to $215.8 million in 2025. | FBI IC3, April 2026 |
| Business email compromise produced 24,768 complaints and $3,046,598,558 in losses, the second-costliest fraud type after investment fraud. | FBI IC3, April 2026 |
| The average amount requested in wire-transfer BEC attacks was $61,732 in Q2 2026, up 45% from $42,663 the quarter before. | APWG / Fortra, August 2026 |
| Email was the most common way US consumers reported being contacted by scammers in 2024, for the second year in a row. | FTC, March 2025 |
| Social media became the costliest fraud contact method in 2025, with $2.1 billion in reported losses. | FTC, April 2026 |
The phishing numbers moved in opposite directions: complaints fell 1% while reported losses tripled. The average reported loss per phishing complaint went from $362 to $1,127. BEC is a different scale again, at $123,005 per complaint, about 109 times the phishing average.
We read the IC3 numbers as a story about targeting. Bulk spam gets fewer victims each year because filters and users have learned, but the attacks that do land are aimed better and ask for more. BEC is spam's expensive cousin: low volume, one well-researched email, a five- or six-figure wire.
Which raises the question of how well the defences hold up.
How well do spam filters work?
Gmail stops more than 99.9% of spam, phishing and malware before it reaches the inbox, according to Google, and authentication rules have cut unauthenticated mail to Gmail users by 75%.
| Statistic | Source |
|---|---|
| Gmail blocks more than 99.9% of spam, phishing and malware, nearly 15 billion messages a day. | Google, October 2023 |
| The number of unauthenticated messages Gmail users receive dropped by 75% after Google started requiring some form of authentication on all mail sent to Gmail addresses. | Google, October 2023 |
| Since February 2024, bulk senders to Gmail must authenticate with SPF, DKIM and DMARC and offer one-click unsubscribe. | Google sender guidelines, 2026 |
| Phishing was the initial access vector in 16% of breaches in the 2026 Data Breach Investigations Report, unchanged from the year before. Vulnerability exploitation led at 31%. | Verizon DBIR, 2026 |
| The human element was present in 62% of breaches, up from 60%. | Verizon DBIR, 2026 |
| In phishing simulations, the median click rate for mobile channels such as voice and SMS is 40% higher than for email. | Verizon DBIR, 2026 |
Filters are very good at the bulk problem and weaker at the targeted one. Gmail's 99.9% describes mass mail; Verizon's 16% describes the few messages crafted well enough to get a person to act.
We build anti-spam tools, and the lesson we keep relearning is that filtering is the last line, not the first. The cheapest spam to stop is the spam that never gets your address. That's why we'd rather hide an address on a page, or put a bot check on a form, than count on a filter to catch every message afterwards.
The same logic applies to the other door spam uses: contact forms.
Does spam also come through website contact forms?
Contact forms are a second entry point for spam, but no public body publishes aggregate contact-form spam numbers; the email statistics above cover messages sent to addresses, not form submissions.
Scripts that crawl pages for addresses can find and fill forms too, and form spam lands in the same inbox through your own website. We're collecting our own data on this, because the numbers people quote are vendor anecdotes. Our contact form spam statistics page tracks what's measurable so far, and the share of internet traffic that comes from bots explains where the senders come from.
For a small site, the practical answer is two steps: don't publish a scrapable address, and don't run a form that a script can fill in a second. If you only do the first, our free email obfuscator takes about a minute.
The original figures in this page come from simple arithmetic on the sources above, shown next.
How we calculated the original numbers
Four numbers on this page are computed here from verified inputs. Each uses one source only, so no figures from different vendors are mixed.
- About 174,000 unwanted emails blocked per second by Gmail. Google's figure of nearly 15 billion a day (October 2023) divided by 86,400 seconds: 15,000,000,000 ÷ 86,400 = 173,611. Because Google says "nearly", treat this as an upper bound for 2023.
- 0.82 spam emails for every legitimate one. Kaspersky's 2025 spam share of 44.99% leaves 55.01% legitimate: 44.99 ÷ 55.01 = 0.818, or about four spam messages for every five legitimate ones. This holds for Kaspersky's telemetry, not for every inbox.
- BEC costs 109 times more per complaint than phishing. From FBI IC3 2025: BEC $3,046,598,558 ÷ 24,768 complaints = $123,005; phishing and spoofing $215,843,126 ÷ 191,561 = $1,127; $123,005 ÷ $1,127 = 109.
- Average phishing loss per complaint tripled. From FBI IC3: 2024 losses $70,013,036 ÷ 193,407 complaints = $362; 2025 = $1,127; $1,127 ÷ $362 = 3.1.
IC3 figures are complaints reported by victims, so all of them undercount real losses. The ratios hold as long as reporting rates are similar across crime types, which IC3 doesn't claim.
What the 2026 numbers say
Spam volume is flat to slightly down; the danger isn't. Kaspersky's spam share fell 2.28 points in 2025, yet malicious attachments rose by 19 million, malware email by 131% (Hornetsecurity) and phishing losses tripled (FBI IC3). Less junk, sharper hooks.
"How many spam emails per day" has no single answer. Gmail's 15 billion is a verified floor for one provider. Totals of 100 billion and more are products of a vendor share and a traffic forecast, and should be cited as estimates with both sources named.
Prevention beats filtering for small sites. Filters stop 99.9% of bulk mail, but a filter only sees what arrives. Keeping an address off scraper lists and keeping bots off forms reduces what arrives in the first place. For the platform side of that, see our WordPress security statistics, and the full set lives on the statistics hub.
FAQ
How many spam emails are sent per day?
Gmail alone blocks nearly 15 billion unwanted emails a day, according to Google (October 2023), so the worldwide total is well above that. No organisation publishes a verified global count. Figures of 100 billion and more come from multiplying a spam share, such as Kaspersky's 44.99%, by a forecast of total email traffic.
What percentage of emails are spam?
44.99% of global email traffic was spam in 2025, according to Kaspersky's annual report, down from 47.27% in 2024. Business-focused filters see less: Hornetsecurity counted 36.9% of business email as unwanted. Cloudflare counted 5.6% as outright malicious.
Which country sends the most spam?
Russia sent 32.50% of the world's spam in 2025, according to Kaspersky, followed by China at 19.10% and the United States at 10.57%. Country of origin reflects where the sending machines sit, which often means hijacked computers in botnets.
How effective are spam filters?
Gmail stops more than 99.9% of spam, phishing and malware before the inbox, according to Google (2023). Targeted mail still gets through: Verizon's 2026 Data Breach Investigations Report found phishing was the entry point in 16% of breaches.
How much money do phishing emails cost victims?
US victims reported $215.8 million in phishing and spoofing losses in 2025, according to the FBI's IC3, up from $70.0 million in 2024. Business email compromise cost $3.05 billion, about $123,005 per complaint, roughly 109 times the phishing average.
When is spam at its highest?
Spam peaks in November and December, according to Kaspersky: 52.87% of all email was spam in November 2025 and 51.80% in December, against about 43.50% for the rest of the year.
Sources
- Kaspersky (Securelist), Spam and phishing in 2025 (February 2026)
- Kaspersky, Kaspersky reports 15% growth in malicious email attacks in 2025 (February 2026)
- Google, New Gmail protections for a safer, less spammy inbox (October 2023)
- Google, Email sender guidelines (accessed November 2026)
- Microsoft, Microsoft Digital Defense Report 2025 (October 2025)
- Cloudflare, Radar 2025 Year in Review (December 2025)
- Hornetsecurity, Cyber Security Report 2025 press release (December 2024)
- Hornetsecurity, Cybersecurity Report 2026 press release (November 2025)
- Proofpoint, The Human Factor 2025 Vol. 2 (August 2025)
- Mimecast, 2025 Global Threat Intelligence Report press release (October 2025)
- Spamhaus, Botnet Threat Update July to December 2025 (January 2026)
- APWG, Phishing Activity Trends Report Q2 2026 (August 2026)
- FBI Internet Crime Complaint Center, 2025 Internet Crime Report (April 2026)
- FTC, New FTC data show a big jump in reported losses to fraud to $12.5 billion in 2024 (March 2025)
- FTC, New FTC data show people have lost billions to social media scams (April 2026)
- Verizon, 2026 Data Breach Investigations Report, Executive Summary (2026)
Methodology. Numbers were collected in November 2026 and checked against the original publisher's page, report or press release, not against other statistics roundups. Vendor figures describe each vendor's own traffic, so shares from different vendors are shown side by side and never combined. We excluded figures we could not open at the source, including the Radicati Group email volume forecast (site unreachable in November 2026) and Statista charts that don't name an original source. Google's 15 billion figure dates from October 2023 and is marked as dated. Last updated 14 November 2026.
Free tool
Keep your own address off spam lists.
The snippet hides your email from harvesters and keeps it clickable for people.