Ethereum is two chains, and most explorers show one
Since the Merge in September 2022, Ethereum has had an execution layer that processes transactions and runs the EVM, and a consensus layer secured by staked validators that decides which blocks are canonical.
Almost every explorer shows the execution layer, because that is where transactions live. Etherscan will tell you everything about a transaction and nothing about the validator that proposed the block containing it, whether that validator has been attesting reliably, or whether anyone was penalised in that epoch.
For most people that is fine. For the roughly one million validators securing the network, and the far larger number of people staking through them, the consensus layer is the one that determines returns — and beaconcha.in is the only thorough view of it.
If you stake, use both. If you do not, you can safely ignore half of Ethereum and most people do.
Internal transactions, and the withdrawal that "did not arrive"
This is the single most common source of confusion on Ethereum, and it produces a steady stream of support tickets at every exchange.
A transaction on Ethereum is a call from one address to another. If the destination is a contract, that contract may call further contracts, and those calls can move value. None of that appears in the original transaction's value field, because it did not happen in the original transaction — it happened during its execution.
Explorers call these internal transactions, and they live on a separate tab. So an address page can show almost no activity while substantial value moved through it, and a transaction can show zero value while transferring a great deal.
When someone reports that a withdrawal has not arrived, the internal transactions tab is the first place to look. The second is the token transfers tab, because ERC-20 movements are log events rather than transaction fields and appear in neither the main list nor the internal one.
| If you are looking for | Look at | Why it is not in the main list |
|---|---|---|
| A plain ETH transfer | Transactions | It is there |
| ETH moved by a contract | Internal transactions | It happened during execution, not as a transaction |
| A token transfer | Token transfers (ERC-20) | It is a log event emitted by the token contract |
| An NFT movement | Token transfers (ERC-721/1155) | Same reason, different standard |
Reading a contract before you sign anything
Etherscan's real moat is not the explorer, it is the verified source registry behind it. Verification means someone submitted the original Solidity along with the exact compiler settings, and Etherscan confirmed it recompiles to the bytecode actually deployed.
That check is reproducible, which is what makes it trustworthy rather than merely authoritative. It has also made "is this verified on Etherscan" a first-order safety question across the entire ecosystem, asked by people who have never written a line of Solidity.
Three things worth knowing when you read one.
Proxies. Most significant contracts are proxies — a permanent address delegating all logic to a separate implementation that can be swapped. A naive explorer shows you forty lines of delegation boilerplate. Etherscan resolves the implementation and offers a "Read as Proxy" toggle; several competitors do not, and the failure is silent.
Owner privileges. Verified does not mean safe. Look for functions only an owner can call: mint without limit, pause transfers, change fees, upgrade the implementation. All of these are legitimate in some designs and all of them are how token rug pulls work.
The write tab. It lets you call state-changing functions directly with your wallet, which is the standard escape hatch when a project's front end goes down. It is also a raw interface with no guard rails — nothing stops you passing the wrong decimals and approving a spend a million times larger than intended.
The privacy position on Ethereum
Ethereum is an account-model chain, which makes it considerably easier to analyse than Bitcoin. Addresses are reused by design — your wallet is one address, and everything you have ever done with it is linked by construction.
That means the usual advice about avoiding address reuse does not apply here; you cannot. What you can control is what you reveal about which address is yours, and explorer lookups are part of that.
Etherscan is closed, commercial and has no onion service. Blockscout is open source, which at least lets you read what the software does, though a hosted instance remains a third party. Otterscan against your own Erigon node is the only complete answer, and the cost is a multi-terabyte archive node.
The honest summary: on Ethereum, privacy from explorers is achievable and privacy from chain analysis is not. Arkham demonstrates the second point more persuasively than any amount of writing, and understanding it is more useful than trying to defeat it.
Where the activity actually went
Much of what used to happen on Ethereum mainnet now happens on rollups, and the explorer question moves with it. Base, Polygon, Arbitrum, Linea and Scroll each have their own explorers, mostly from the Etherscan family, and one V2 API key covers all of them.
The thing that changes is what "confirmed" means. On mainnet a transaction in a block is settled. On a rollup it is confirmed by a sequencer, and settlement on Ethereum happens later — minutes to hours for a validity rollup, days for an optimistic one with its challenge window.
Both explorers use the word "confirmed". Only one of them means finality in the strongest sense, and the field that distinguishes them is the L1 batch status. If you move significant value across rollups, that is the field to learn.
Our reviews of BaseScan, Arbiscan and LineaScan cover how each one presents it, and the differences between optimistic and validity rollups that make the timelines so different.