Ethereum is a public network that records balances and also runs programs. That second part is what separates it from a network built mainly to move one asset. Understanding accounts, gas, and contracts explains most of what you will see in a wallet, and it explains why a transfer can fail in ways a bank transfer never does.
Ethereum runs programs as well as recording balances. That flexibility is why so much is built on it, and why a transaction can do something you did not intend in ways a simple transfer cannot.
- How key-controlled accounts differ from contract accounts.
- Why gas is charged even when a transaction fails.
- What a token approval actually grants.
- Why the network you use matters as much as the address.
What Ethereum is trying to be
Ethereum keeps a shared record of balances, like many blockchains, but it also stores and executes code. Anyone can publish a program to the network, and anyone can send a transaction that runs it. The result is recorded for everyone to check. That design is why so many applications, from token issuance to marketplaces, were built on it rather than on a network that only transfers a single asset.
This flexibility is not automatically an advantage. More capability means more ways for a transaction to do something you did not intend, and more places where a mistake becomes permanent. Treat Ethereum as general-purpose infrastructure whose behavior depends entirely on the specific code you interact with, not as a product with one guaranteed experience.
Accounts, keys, and who can act
Two kinds of account exist. One is controlled by a private key held in a wallet; it can start transactions. The other is a contract account, which holds code and only acts when something calls it. Your wallet address is the first kind. The address of a token or an application is usually the second, which is why sending assets to a contract that does not expect them can strand those assets.
An address does not tell you what sits behind it. Wallet software often labels known contracts, but a label is a convenience added by the interface, not a fact stored on the network. Before sending anything to an unfamiliar address, confirm from an independent source what that address is and whether it is designed to receive the asset you are sending.
Two kinds of account
| Key-controlled account | Contract account | |
|---|---|---|
| Controlled by | A private key in your wallet. | Its own code, which runs when called. |
| Can start a transaction | Yes. | No; it only acts when something calls it. |
| Typical example | Your wallet address. | A token, a marketplace, an application. |
| Risk if you send to it | Standard transfer risk. | Assets can be stranded if the code does not expect them. |
Smart contracts are programs, not promises
A smart contract executes its own instructions consistently. That reliability is often described as trustlessness, which is easy to misread. The code will do what it says; it will not do what a marketing page says, and it cannot know that its author made an error. Audits reduce some risk and do not eliminate it. Code can also be upgradeable, meaning its behavior may change under rules you have not read.
When an application asks your wallet to sign something, you are approving a specific interaction with specific code. Wallets increasingly try to explain what a request will do, and those explanations vary in quality. If a request is unclear, the safe response is to reject it and investigate rather than to approve and hope the description was complete.
A practical consequence is that reading a contract’s reputation is not the same as reading its code. Most people cannot audit a contract, and that is a reasonable position to be in. What is achievable is checking who deployed it, whether it is upgradeable, how long it has operated, and what independent parties have said about it, then sizing your exposure to reflect the fact that you are relying on others’ judgement.
Ether pays for computation
Every transaction consumes gas, and gas is paid in ether. A simple transfer uses a small, predictable amount. Calling a contract can use far more, because the network is executing instructions and storing results. The total cost combines the work required with a fee rate that changes as demand for block space changes, so identical actions can cost noticeably different amounts at different times.
Two practical consequences follow. First, you need ether in the account that sends a transaction, even if you are moving a token that is not ether. Second, a transaction can fail after being included, and the gas spent on the failed attempt is generally not returned. Wallets estimate costs before you confirm; treat the estimate as an estimate.
Tokens live on top of the network
Most assets you encounter on Ethereum are tokens defined by contracts rather than separate networks. Shared standards let a wallet display many tokens without custom code for each one. That is convenient, and it also means creating a token is easy. A familiar name or logo in a wallet list proves nothing about who created it, what it does, or whether anyone will buy it.
Identify a token by its contract address from an official source, not by its displayed symbol. Duplicate and lookalike tokens are common, and some are deliberately designed to be received but not sold. Unexpected tokens appearing in a wallet are usually unsolicited, and interacting with them is the risk, so investigate before you attempt to trade or approve anything.
Unexpected tokens arriving in a wallet are common and are best left alone. Sending a token to an address costs the sender very little, and unsolicited arrivals are frequently bait: attempting to trade or approve them is the point at which something goes wrong. Receiving a token is harmless; interacting with one you did not seek out is where the risk begins.
Upgrades, validators, and staking
Ethereum uses proof of stake. Validators deposit ether, propose and attest to blocks, and can lose part of their deposit for provable misbehavior or extended downtime. Running a validator has technical and capital requirements. Pooled and custodial services lower that barrier and add their own operator, contract, withdrawal, and regional considerations, which is a different set of trade-offs rather than a simpler version of the same one.
The network also changes over time through coordinated upgrades. Fee mechanics, data availability, and validator rules have all changed before and can change again. Documentation you read a year ago may describe behavior accurately for that period and not for today, so check the date on anything technical you rely on.
An approval permits a contract to move a specified token from your account under set conditions. It stays in place after the action that prompted it, sometimes without limit and sometimes indefinitely.
Reviewing and reducing old approvals is ordinary maintenance rather than an advanced technique. A permission granted to an application you last used a year ago is still a live permission today.
Layer 2 networks change the address you use
Much activity has moved to networks that settle to Ethereum while processing transactions separately. They usually offer lower fees, and they are distinct networks with their own bridges, fee assets, confirmation behavior, and risks. Your wallet address may look identical across several of them, which is precisely why network selection errors are so easy to make.
Before any transfer, confirm which network the receiving service expects, character for character, and confirm that the specific token version is supported there. An exchange that supports a token on the main network may not support the same token bridged to another. The address matching is not the check that matters; the network and asset pairing is.
Withdrawing from a layer 2 network back to the main network is worth understanding before you deposit into one. Depending on the design, it can involve a waiting period measured in days, a third-party liquidity provider charging for immediacy, or a process that behaves differently under congestion. The route in is almost always faster and simpler than the route out.
How to study Ethereum without risking funds
You can learn most of this without spending anything. Read a public transaction on an explorer and identify the sender, the recipient, the asset, the fee, and whether a contract was called. Read the documentation for one token standard until the transfer and approval functions make sense. Follow a wallet’s confirmation screen for a hypothetical action and write down what it says it will authorize.
When you do act, start with an amount whose loss would not matter, keep the network and asset consistent, and record the transaction hash. The goal is not to memorize a specific interface but to be able to describe, in your own words, what you authorized and where the result was recorded. If you cannot describe it, that is the signal to stop.
Key terms to keep handy
- Ether
- The network’s native asset, used to pay for transactions and to secure the network through staking.
- Account
- An entry on the network that can hold a balance; some are controlled by keys, others by contract code.
- Smart contract
- Program code stored on the network that runs when a transaction calls it.
- Gas
- The unit that measures the computational work a transaction requires.
- Token standard
- A shared set of rules that lets wallets and applications handle many different tokens the same way.
- Validator
- A participant that proposes and attests to blocks under the network’s proof-of-stake rules.
Find the network selector
The Binance buy cryptocurrency guide shows a purchase on a large platform. The screen worth studying is the one that asks which network to use, because that single choice causes more permanent losses than any other field.
The same asset can exist on several networks with an identical-looking address. Matching the address is not the check that protects you; matching the network and the asset version is.
Sources and further reading
- Ethereum accounts documentation (opens in a new tab)
- Ethereum smart contracts overview (opens in a new tab)
- Ethereum gas and fees (opens in a new tab)
Frequently asked questions
Is Ethereum the same as ether?
No. Ethereum is the network. Ether is the asset used to pay transaction costs and to secure the network through staking. People often use the names interchangeably in conversation, but wallets, exchanges, and documentation treat them as separate things.
Why do I need ether to move a token?
Transaction costs on Ethereum are paid in ether regardless of what the transaction moves. If an account holds a token but no ether, it generally cannot send that token until the fee asset is available in the same account on the same network.
Can a smart contract be changed after it is published?
Some contracts are designed to be upgradeable through defined mechanisms, and others are not. The distinction matters because upgradeability introduces a party who can alter behavior. Check the specific contract’s documentation rather than assuming either case.
What is a token approval?
An approval is a transaction that permits a contract to move a specified token from your account under set conditions. Approvals persist until changed, so reviewing and reducing old approvals is a normal maintenance task rather than a one-time action.
Is a layer 2 network safer or riskier than Ethereum?
It is different. Layer 2 networks generally reduce fees and introduce their own bridge, sequencer, and withdrawal considerations. Neither answer applies universally, so read the specific network’s documentation about how withdrawals and failures are handled.
What should I read next?
Read the guides to network fees and crypto wallets. Fees explain why the same action costs different amounts, and wallets explain who controls the keys that authorize any Ethereum transaction.
Crypto can lose substantial value, and transfers may be irreversible. This guide is educational, not financial, legal, or tax advice. Exchange access and features depend on your location.
Independent educational writing. How we research and correct our guides.