What Is a Smart Contract? How They Work & Examples

What is a smart contract?

So, what is a smart contract? It is a computer program that lives on a blockchain and runs exactly as written. The name can mislead, because a smart contract is not artificial intelligence and not always a legal contract. It is code that follows simple "if this happens, then do that" rules.

The idea comes from computer scientist Nick Szabo, who described smart contracts in the 1990s as promises written in digital form, together with the protocols that carry them out. His favourite example was the vending machine: you insert the right coin, press a button, and the machine releases the product without a shopkeeper. Blockchains made the idea practical because they offer a shared record that no single party controls. Ethereum, which launched on 30 July 2015, brought smart contracts to a wide audience.

How smart contracts work

Most of what follows uses Ethereum as the main example, because its design is well documented. Other blockchains work in similar but not identical ways.

Code and state

A smart contract is a collection of code (its functions) and data (its state) stored at a specific address on a blockchain. Developers write it in a high-level language, compile it, and deploy it with a transaction. Once deployed, anyone can usually read the public code and call its functions.

Triggers

A smart contract does not run on its own schedule. It waits until a wallet or another contract sends it a transaction. A contract also cannot look up outside information, such as a share price or the weather, unless a separate data service called an oracle supplies it.

Execution

When a transaction arrives, the nodes of the network run the same code and must reach the same result. If the conditions in the code are met, the contract updates its state, for example by moving tokens. If they are not met, the transaction fails and its changes are undone. The outcome is then recorded on the blockchain.

Gas

On Ethereum, each step of a contract's execution costs a small amount of computing effort, measured in gas and paid in Ether (ETH). Gas (Gas Fees & Gas Limit) stops programs from running forever and pays for the network's work. Deploying a contract costs far more than a simple transfer. Other blockchains use different fee systems.

The contract lifecycle

1. Write: a developer writes the contract code and its rules.
2. Test and review: the code is tested and, ideally, reviewed by independent specialists.
3. Deploy: the compiled code is sent to the blockchain in a transaction and receives its own address.
4. Call: a user or another contract sends a transaction that triggers one of its functions.
5. Verify and execute: network nodes run the code and check the conditions.
6. Record: if the conditions are met, the new state is written to the blockchain and assets move. If not, the transaction is rejected.
7. Persist: the contract stays on the blockchain. By default it cannot be deleted, and past interactions cannot be reversed.

As an example, imagine a simple escrow contract. A buyer sends payment to the contract. If the seller's delivery is confirmed before a deadline, it releases the payment to the seller, and if not, it returns the payment to the buyer. No third party holds the money, but the delivery confirmation still has to come from somewhere, usually an oracle or a signed approval.

Smart contracts vs traditional contracts

A traditional contract is an agreement written in human language and enforced by courts, arbitration or the parties' goodwill. A smart contract is enforced by software. The two can overlap, but they are not the same thing.

Many people ask whether a smart contract is legally binding. There is no single global answer. In November 2021, the Law Commission of England and Wales concluded that its existing legal framework can already support smart legal contracts, but that is one jurisdiction's view.

Platforms that run smart contracts

Several blockchains support smart contracts. Their languages, fees and designs differ, and this list changes over time, so treat it as a snapshot of well-known examples as of 2026 rather than a complete or ranked list.

Because EVM-compatible chains share the same tools, a contract written for Ethereum can often be deployed on them with few changes. Always check how each network is run, how it charges fees and what its security record looks like before you use it.

Use cases

Smart contracts are the building blocks of most on-chain applications. A dApp (Decentralized Application) usually pairs a website with one or more contracts that hold the rules and the assets. Common areas include the following.

• DeFi: lending pools, token swaps and stablecoins use contracts to hold funds and apply rules for borrowing, trading and interest without a company running the book.
• NFTs: a token contract records who owns each item and the rules for transferring it. The image or file itself is often stored elsewhere.
• DAOs: a DAO (Decentralized Autonomous Organization) puts voting rules and a shared treasury into contracts, so members' votes can trigger payments. The first well-known example, called The DAO, was launched in 2016.
• Insurance: parametric policies could pay out automatically when a trusted data feed reports an agreed event, such as a flight delay. Most remain experimental.
• Supply chain: contracts can record handovers and release payments when a shipment is confirmed, but they depend on reliable real-world data entering the chain.

Many real-world use cases still depend on oracles, sensors or human input. A smart contract can be exact about what it is told, but it cannot check on its own whether the information is true.

Benefits and limitations

Benefits

• Automation: payments and transfers happen when the conditions are met, without manual approval.
• Transparency: on public blockchains, anyone can inspect the code and the history of its use.
• Always on: a deployed contract can be used at any time, without office hours.
• Less reliance on one party: no single company controls how the contract executes, although developers or administrators may still hold special powers.
• Composability: contracts are public, so one can call another, like open building blocks.

Limitations

• No built-in outside knowledge: contracts cannot see off-chain data without an oracle.
• Fees: every interaction costs gas or a similar fee, and deployment can be expensive.
• Rigidity: code cannot handle the vague or changing situations that human agreements often do.
• Size limits: on Ethereum, a single contract has a maximum size of 24KB.
• Irreversibility: a transaction sent by mistake usually cannot be undone.
• Legal uncertainty: how courts treat disputes over code varies by country.

Risks

Smart contract risk is mostly about code, data and trust. Money held in a contract can be lost quickly if something goes wrong, and there is usually no help desk to reverse it.

Bugs and exploits

Contracts are written by people, and people make mistakes. An attacker who finds a flaw can sometimes drain funds in a single transaction. See Exploit (Crypto Exploits & Hacks) for how such attacks are described and tracked.

Immutability

Public blockchains make deployed code hard to change. That protects users from sudden edits, but it also means a flaw can stay live. Some contracts are built to be upgradeable, which fixes bugs but means an admin key or vote can alter the rules, so the trust moves to whoever holds that power.

Oracle risk

If a contract relies on a price or event feed, a wrong or manipulated feed can cause wrong payouts. The contract can be flawless and still produce a bad result.

Smart contract languages and audits

The language depends on the blockchain. Solidity is the most widely used language on Ethereum and EVM-compatible chains. Vyper is a Python-like alternative on Ethereum. Solana programs are commonly written in Rust, while Aptos and Sui use Move, and code is compiled before deployment.

Because mistakes are costly, projects often pay for a Smart Contract Audit. In an audit, security specialists read the code, run tests and automated tools, and report problems for the team to fix. Some projects also run bug bounty programs or use formal verification, which checks code against mathematical rules.

• An audit is a snapshot. Later code changes may not be covered.
• Check who did the audit, which version was reviewed, and whether the findings were fixed.
• Even a clean report cannot guarantee that a contract is free of bugs.

Frequently asked questions

Related terms

• Smart Contract Audit
• Solidity
• Ether (ETH)
• Gas (Gas Fees & Gas Limit)
• DAO (Decentralized Autonomous Organization)
• dApp (Decentralized Application)
• Nick Szabo
• Exploit (Crypto Exploits & Hacks)

Sources

1. Smart contracts ethereum.org
2. History and Forks of Ethereum ethereum.org
3. Smart Contracts: Building Blocks for Digital Markets Nick Szabo (1996, archived copy)
4. Smart legal contracts Law Commission of England and Wales
5. The DAO hack at 10: From $50 million exploit to a $130 million Ethereum security fund The Block
6. Programs Solana documentation