---
title: "Data Import"
space: "Frappe Framework"
url: "https://www.prilk.com/docs/frappe-framework/data-import"
updated: "2026-05-14"
---

## Data Import

Bulk-loading data via CSV/XLSX is one of Frappe's strengths — used for initial migration, ongoing customer/item updates, or routine bulk operations.

### The Data Import DocType
Open **Data Import → New**. Pick:
1. DocType (e.g., Customer, Item, Sales Invoice)
2. Action — **Insert New Records** or **Update Existing Records**
3. Upload a CSV/XLSX file
4. Map columns to fields (auto-detected when headers match field labels)
5. Hit **Start Import** — runs in the background, you get a report when done

### Template download
For a fresh import, click **Download Template** on the Data Import form to get a CSV with the exact column headers and field types pre-filled. Easier than guessing field names.

### Child tables
For DocTypes with child tables (Sales Invoice with its items), the import format flattens them: one row per parent, repeating the parent name in subsequent rows for each child line. The template explains the convention.

### Updates by name
To update existing records, include the `ID` column with the record name. Frappe matches and updates. Skipped rows are reported in the import log.

### Validation
Each row is validated against the DocType's rules — mandatory fields, link validations, custom validate hooks. Failed rows are listed in the log with the reason; valid rows still import.

### When to use the API instead
Data Import is interactive (one DocType at a time). For nightly syncs from another system, use the [API](/docs/frappe-ecosystem/api) — it's faster, scriptable, and avoids the desk UI overhead.

### Large imports
For 100k+ rows, prefer chunked imports or direct DB scripts (run via `bench --site mysite execute`). The Data Import UI works but processes one record at a time, which is slow at scale.