A distributed ledger is a shared record of data that lives on many computers at once. Each computer, called a node, keeps a copy, and the network keeps those copies in sync so everyone sees the same state without a central administrator. Blockchains are one kind of distributed ledger, but not the only kind.
Distributed ledgers rely on peer-to-peer networking and a way for nodes to agree on valid updates, known as consensus. When a new entry is proposed, nodes verify it and update their local copy only if the proposal passes the rules. The result is a single, consistent history shared across the network. Modern implementations use cryptography and fast messaging between nodes to keep replicas aligned within seconds.
Many distributed ledgers aim to make past entries tamper-evident. In blockchains, entries are grouped into blocks that reference prior blocks, which makes changing history extremely hard once the network accepts it. Other distributed ledgers may store entries without blocks while still replicating the same data across nodes.
Distributed ledger technology (DLT) is the umbrella term. Blockchain is a specific design that chains blocks in order. Other designs exist, such as directed acyclic graphs, which also replicate data across nodes. Treat blockchain as a subset of DLT rather than a synonym.
Distributed ledgers can be arranged by who may read and write: public or private, and by how participants are admitted: permissionless or permissioned. Real systems mix these choices, for example, private-permissioned networks for a group of firms, or public-permissionless networks that anyone can join.
Beyond cryptocurrencies, distributed ledgers show up in payments, asset tracking, supply chain tracing, digital identity, and public registries. The common thread is a shared, verifiable record that different parties can rely on without routing every change through one database owner.
The architecture can be harder to design and maintain than a classic central database. Some networks struggle to scale under heavy use. Energy use varies by consensus method, and proof-of-work systems in particular can consume substantial power.