How does the converter decide the column headers?

It takes the union of every key across all objects in the array, then sorts the headers alphabetically. So if the first object has name and age and a later one adds a phone field, phone becomes its own column and earlier rows simply leave that cell blank. Nothing is dropped because one object happens to be missing a key.

What happens to a scalar array like tags: ["admin", "beta"]?

A list of simple values (strings, numbers, booleans) is joined into a single comma-separated cell - tags becomes the cell "admin, beta". The row count never changes, so one object in your array is always exactly one row in the sheet.

Will IDs like U-001 or a zip of 00100 stay exact?

Yes. Values that are quoted strings in your JSON stay text, so leading zeros survive (00100 does not become 100) and order codes like U-001 are untouched. Long numeric IDs are read with a big-number-safe parser, so they keep every digit instead of being rounded or shown in scientific notation.

What if some objects in the array hold a nested list of objects?

A field whose value is a list of objects (for example a line_items array inside each order) is kept as JSON text in one cell by default, so each top-level object stays a single row. If you want those expanded instead, see the nested JSON guide for how deeper structures are handled.

Convert a JSON Array to Excel Columns (Step-by-Step)

Last updated: 13 June 2026

The most common JSON shape is an array of objects - a list where each item has the same kind of fields. That maps perfectly onto a spreadsheet: each object becomes a row and each field becomes a column. This guide walks through one realistic array end to end so you can predict exactly what columns and rows you'll get before you even open the file.

A realistic array of objects

Here is a small users array that packs the three cases you hit in real data - a scalar list (tags), a nested object (address), and a leading-zero code:

[
  {
    "id": "U-001",
    "name": "Ada Lovelace",
    "tags": ["admin", "beta"],
    "address": { "city": "London", "zip": "SW1A" }
  },
  {
    "id": "U-002",
    "name": "Linus Torvalds",
    "tags": ["maintainer"],
    "address": { "city": "Helsinki", "zip": "00100" }
  }
]

The resulting table

In the default Separate columns mode that array becomes two rows with these headers (the converter sorts the column set alphabetically):

address.city	address.zip	id	name	tags
London	SW1A	U-001	Ada Lovelace	admin, beta
Helsinki	00100	U-002	Linus Torvalds	maintainer

Nested objects flatten to dotted columns

The address object did not land in a single cell - it spread into address.city and address.zip. That dotted-path naming is what lets you filter and pivot on a sub-field in Excel. If you would rather keep a subtree intact, switch to JSON in one cell and address stays as one JSON string. For deeper trees, see the nested JSON to Excel guide.

Scalar arrays become one comma-joined cell

The tags field held a list of plain strings, so it collapsed into a single cell - admin, beta for the first row and maintainer for the second. This keeps each object on exactly one row instead of multiplying rows out for every tag. A list of objects (rather than strings) is instead kept as JSON text in one cell, again so your row count stays one-to-one with the array.

Different keys across objects

Real arrays are rarely uniform. If one object carries a phone field and the others don't, the converter still builds the column from the union of all keys and simply leaves the cell blank where a value is missing - so a sparse field never silently drops the rest of the row.

Convert it

Paste your JSON array, or drop a .json file (up to 15 MB), into the converter.
Preview the columns and rows to confirm the mapping.
Download the .xlsx and open it anywhere.

It's free, needs no signup, and runs entirely in your browser. If your data came from an API wrapper such as { "data": [ … ] } rather than a bare array, the API response to Excel guide shows how the record array is detected automatically.

Frequently asked questions

How does the converter decide the column headers?: It takes the union of every key across all objects in the array, then sorts the headers alphabetically. So if the first object has name and age and a later one adds a phone field, phone becomes its own column and earlier rows simply leave that cell blank. Nothing is dropped because one object happens to be missing a key.
What happens to a scalar array like tags: ["admin", "beta"]?: A list of simple values (strings, numbers, booleans) is joined into a single comma-separated cell - tags becomes the cell "admin, beta". The row count never changes, so one object in your array is always exactly one row in the sheet.
Will IDs like U-001 or a zip of 00100 stay exact?: Yes. Values that are quoted strings in your JSON stay text, so leading zeros survive (00100 does not become 100) and order codes like U-001 are untouched. Long numeric IDs are read with a big-number-safe parser, so they keep every digit instead of being rounded or shown in scientific notation.
What if some objects in the array hold a nested list of objects?: A field whose value is a list of objects (for example a line_items array inside each order) is kept as JSON text in one cell by default, so each top-level object stays a single row. If you want those expanded instead, see the nested JSON guide for how deeper structures are handled.

Related guides

See all guides or the FAQ.

Convert your JSON array to Excel →