Bitcoin recorded 686,939 active addresses in the latest 24-hour window on Glassnode’s live chart. It is a substantial measure of successful on-chain participation, but it is not a count of Bitcoin users—and it cannot, by itself, establish whether adoption is rising or falling.
The distinction is more than semantic. A single user can activate multiple addresses through ordinary wallet behavior, while an exchange or custodian can represent many users through relatively few on-chain addresses. Meanwhile, some forms of bitcoin exposure take place without a new on-chain transaction at all. The result is a metric that is useful, immediate and easy to overread.
686,939 active addresses is a count of ledger participants, not Bitcoin users
Glassnode defines an active Bitcoin address as a unique address that sent or received funds in a successful transaction. On that definition, the 686,939 reading describes the number of distinct addresses involved in successful transfers during the chart’s latest available 24-hour period.
That makes the series a measure of address-level breadth within Bitcoin’s ledger activity. A higher reading can show that more unique addresses were involved in transactions over the selected period; a lower one shows the reverse. It does not reveal how many natural persons, businesses or investment accounts stand behind those addresses.
Coin Metrics uses a closely related definition: unique addresses participating in a ledger change during a specified trailing window. Its documentation explicitly calls active addresses a proxy for users and cautions that the measure can inherit distortions specific to a blockchain. The proxy can be informative without becoming an identity count.
Time window matters, too. A daily active-address reading captures participation over a day, rather than a permanent installed base. Comparing it with another daily reading may help describe changes in on-chain activity. Calling either number “Bitcoin adoption” adds claims about people, ownership and economic use that the raw address count does not contain.
Bitcoin’s address architecture can inflate activity while custody can compress it
A raw address total does not map cleanly onto people. Wallets commonly generate a fresh receiving address for each transaction, so one person or organization can create and activate many addresses. Bitcoin Core documentation warns that address reuse damages privacy, making new receiving addresses consistent with avoiding that reuse.
The result, as Glassnode notes in its on-chain activity guide, is that more active addresses need not mean proportionately more people using Bitcoin. The same participant may appear as several active addresses across transactions or during one analytical period. At the same time, an exchange or custodian address can combine balances and transactions associated with many customers, making a raw count understate people with an economic interest in bitcoin or intermediary access to it.
Glassnode’s response is to offer entity-adjusted metrics that cluster addresses believed to belong to the same entity. That is a different unit, not a headcount: the provider says entity counts should not be treated as counts of individuals. An entity may be an exchange, service, institution or another grouping rather than one person.
Address generation and pooled custody pull the raw measure in opposite directions. It can overstate individual participation in the first case and understate it in the second; neither direction is fixed, because the balance can change with wallet practices and movements among custody arrangements.
A stable balance-holding address base challenged headline activity during 2022–23
The historical comparison drawn by Coinbase Institutional illustrates the gap between transactional activity and a different kind of address base. Its research found that daily active Bitcoin addresses with a nonzero end-of-day balance remained around 400,000 from the beginning of 2022, even as headline active-address counts, transaction counts, bitcoin’s price and fees changed and Ordinals emerged.
That finding does not make balance-holding addresses a definitive user metric. They remain addresses, with the same broad identity and custody limitations. But it does show that the conclusion changes when the question changes—from which addresses participated in transfers to which active addresses finished the day with bitcoin on them.
The Coinbase Institutional analysis is particularly useful because the headline measures did not move in a vacuum. Transaction counts and fees can respond to demand for block space. Price can move with market conditions. Ordinals introduced activity that need not map neatly onto a broader population of bitcoin holders. A steady balance-holding series amid those shifts complicates the premise that a rise in active addresses directly measures growth in adoption.
For interpretation, this means active addresses are best read as an activity series with a defined construction, not as a catch-all referendum on Bitcoin’s user base. The question is not whether the number is “real.” It is whether its change reflects more distinct economic participants, more address turnover, different transaction behavior, or a mixture of all three.
Institutional exposure and inscription traffic separate adoption signals from address counts
In the second quarter of 2025, Fidelity Digital Assets reported declines in active addresses, new addresses and transaction counts. Yet addresses holding at least $1,000 of bitcoin increased 16.2% quarter over quarter, reaching 12,972,926. Fidelity identified spot bitcoin ETPs, public-company proxies, long-term holders and institutional custody as routes that can raise bitcoin exposure without raising on-chain address activity. The Q2 figures therefore show that growing exposure and a weakening active-address series can coexist.
Fidelity’s Q3 2025 report supplies a separate example: monthly transaction counts rose 50%, active addresses fell 1.7% and new addresses rose 3.1%. Fidelity attributed part of the separation to low-fee inscription, BRC-20 and Runes activity.
These measures describe different aspects of on-chain use. Transaction intensity can rise without a comparable expansion in address breadth, so transaction counts and active-address counts should be read together, not interchangeably.
Use active addresses alongside balances, flows and transaction composition
Active addresses remain a practical on-chain measure. They provide a direct count of unique addresses that successfully sent or received funds during a chosen period, and their consistent definition makes changes over time worth examining. But the metric’s strongest claim is narrow: it measures the breadth of address participation in recorded on-chain activity.
A more resilient reading puts it beside metrics built to answer different questions. Balance-holding address measures can indicate whether addresses retained bitcoin at the end of a period. Entity-adjusted data can reduce some distortion from address fragmentation, while retaining the warning that entities are not individuals. Transaction counts and fee conditions can help identify whether a shift is tied to heavier block-space use rather than a broader participant base.
Exposure measures require still another lens. Fidelity’s second-quarter data show why on-chain activity may miss participation through spot bitcoin ETPs, public-company proxies and institutional custody. Conversely, the third-quarter figures show why a transaction surge deserves scrutiny of what is driving the transactions before it is framed as a user surge.
The 686,939 active-address reading is therefore meaningful as a description of one day’s ledger activity. It becomes a claim about adoption only after the analyst confronts the address-generation behavior, custody arrangements, balance distribution and transaction types that the headline number leaves unresolved.
