What happens in Excel when a CSV has more than 1,048,576 rows?

An Excel worksheet can only display up to 1,048,576 rows × 16,384 columns (in xlsx format; the older xls format stops at 65,536 rows). If you open a CSV that exceeds this, Excel loads rows only up to the limit and the rest never appears on screen — and if you save the file in that state, the overflow can be silently dropped. The data is still intact in the original CSV file, so the safe approach is to never open-and-save a large file in Excel. Instead, use an extraction tool to pull just the rows and columns you need directly from the original.

I opened a CSV and the text looks garbled. How do I fix it?

This happens when the file's character encoding doesn't match the way the program interprets it. Files exported from Korean systems, for example, mix UTF-8 and EUC-KR encodings, so opening one with the wrong assumption turns the text into gibberish. Upload the file to a tool that auto-detects encoding, or manually switch to EUC-KR (Korean Excel) when the preview looks broken, and the text displays correctly. The data itself is not damaged — you only need to change how it's read, not the original file.

Can I extract just specific columns from a CSV with hundreds of thousands of rows?

Yes. Upload the file to a large-CSV extraction tool, switch to column-extraction mode, and select the columns you need (multiple columns can be selected at once) to download a result containing only those columns. If you need only rows matching a condition, use row-matching mode and specify the condition column and value. Choose a tool that processes everything inside your browser without uploading the file to a server, and you can safely handle customer and sales data too.

When a 100MB CSV Won't Open in Excel — Don't Open It, Extract From It

Jun 15, 2026

You've been here, right? You double-click a CSV exported from your settlement system, and Excel just spins with a busy cursor until it hits "Not Responding." Once a file passes 100MB, that's completely routine. Even if it finally opens after a few minutes, the pain isn't over: one Ctrl+F freezes everything again, dozens of columns force you to scroll sideways forever — when all you actually need is three or four columns and a few thousand rows that match a condition.

Here's the short answer: files like this should never be "opened" in Excel — they should be "extracted" from. Excel is built to load an entire file onto your screen, so the bigger the file, the slower it gets. Extraction works the other way around: it pulls out only the matching rows or the columns you pick and hands them back as a new file, so no matter how big the original is, the result stays light. No Power Query, no Python — it works in your browser, no coding required.

A huge CSV isn't something you open — it's something you extract from. Pull just the matching rows and needed columns, and even a 100MB file becomes a lightweight result.

Let's diagnose the symptom first — a cause-by-symptom table

Problems with large CSVs almost always fall into one of these three. Find which one is yours first.

Symptom	Cause	Fix
Won't open, or freezes for ages ("Not Responding")	Excel tries to load the entire file into memory — at tens to hundreds of MB, just loading it is a struggle	Don't open it — extract only the rows and columns you need
Data gets cut off partway	Excel's worksheet limit — at most 1,048,576 rows × 16,384 columns (the older xls format: 65,536 rows). Anything beyond never loads	Extract directly from the original CSV (never save it from Excel)
Text turns into gibberish	Character-encoding mismatch — files from Korean systems, for instance, mix UTF-8 and EUC-KR, and opening with the wrong one garbles the text	Use encoding auto-detection, or manually select EUC-KR

Excel xls (legacy)

65,536 rows

Excel xlsx

1,048,576 rows

Excel's maximum worksheet rows — a CSV itself is plain text with no row limit, so files that exceed the cap are common in real work

Open a million-plus-row CSV in Excel, hit Save — and data disappears

When a file exceeds the limit, Excel loads only up to row 1,048,576 and the rest never appears on screen. Save in that state, and you can end up with a file missing everything that was cut off. The original CSV still holds all the data — so never "open and save" a big file; extract from the original instead.

Extracting without code — done in 3 steps

Search for solutions and you'll find Power Query (Excel's data-transformation feature) or Python scripts — both real time investments to learn. If it's urgent work, the fastest route is uploading the file to Large CSV Data Extract. It handles files up to 250MB on desktop, and the file never leaves your machine — everything runs inside your browser.

STEP 1

Upload the file

Auto-detects encoding · no garbled text

STEP 2

Pick an extraction mode

Matching rows, or just the columns you need

STEP 3

Preview → download

Check the result, save as a light file

The whole process runs inside your browser — the file is never uploaded to a server

USE THIS WHEN

"I only need rows containing this value" → matching-row extraction

Specify a condition column and a value, and it pulls only the rows that match. For example: only the rows with a certain product code from an order export, or only one merchant's transactions from a settlement file. Leave the value empty to target all rows — which also makes it a quick way to just count how many rows there are.

USE THIS WHEN

"Too many columns — I just want a few" → column-value extraction

Pick just the columns you need out of dozens (multiple selection supported) and download a file containing only those columns — say, order number, amount, and date to paste into a report. It can also pull de-duplicated unique values, handy for figuring out "what values does this column actually contain?"

Garbled text is fixed by simply picking the right encoding

The tool auto-detects the encoding when you upload, and if the preview still looks broken, just switch to EUC-KR (Korean Excel) manually. The data was never damaged — it was only being read the wrong way. Leave the original alone, change how it's interpreted, and everything displays fine.

People who hit this problem every single day

Online sellers & e-commerce operators — monthly order and settlement exports from marketplaces easily run tens of MB. Perfect for pulling just one product's or one period's rows for reconciliation.
Marketers — extracting campaign rows, or just the email and conversion columns, from raw ad-performance reports and customer lists to build a report.
Accounting & settlement staff — picking out one merchant's or one approval type's transactions from card-processor files to cross-check. Browser-only processing matters here: sensitive revenue data never leaves your machine.
Open-data users — nationwide public datasets can run to millions of rows; extract just your region's rows and the file shrinks dramatically.
Back-office & CS operators — pulling members matching a specific condition from member and usage-log exports for follow-up.

3 things worth knowing

Never upload a CSV full of customer data to just any online tool

Order and member data usually contain personal information — names, phone numbers, addresses. If a service uploads your file to a server for processing, that data is leaving your control. Make it a habit to check whether processing is fully client-side (in-browser) first. The extraction tool introduced here never sends your file outside the browser.

On mobile, the limit is 30MB — use a desktop for big files

Mobile browsers have far less memory to work with, so the processing cap is 30MB there (250MB on desktop). For files beyond a few dozen MB, desktop is both safer and faster. Files over a million rows still process, but you'll see a notice that the preview may slow down.

Wrap-up — don't fight the big file; take only what you need

Facing a huge CSV, what we ever actually wanted was never "to open the file" — it was "those few rows and columns inside it." Upload the file to Large CSV Data Extract, specify a condition or the columns, and what used to be minutes of Excel freezing becomes a few seconds of extraction. Free, no install, and the file never leaves your browser.

For the exact source on Excel's limits, see Microsoft's official documentation: Excel specifications and limits and What to do if a data set is too large for the Excel grid.

Excel's limit is 1,048,576 rows. For data beyond it, extract — don't open. That's the fastest and safest way to handle a large CSV.

Frequently asked questions

: An Excel worksheet can only display up to 1,048,576 rows × 16,384 columns (in xlsx format; the older xls format stops at 65,536 rows). If you open a CSV that exceeds this, Excel loads rows only up to the limit and the rest never appears on screen — and if you save the file in that state, the overflow can be silently dropped. The data is still intact in the original CSV file, so the safe approach is to never open-and-save a large file in Excel. Instead, use an extraction tool to pull just the rows and columns you need directly from the original.
: This happens when the file's character encoding doesn't match the way the program interprets it. Files exported from Korean systems, for example, mix UTF-8 and EUC-KR encodings, so opening one with the wrong assumption turns the text into gibberish. Upload the file to a tool that auto-detects encoding, or manually switch to EUC-KR (Korean Excel) when the preview looks broken, and the text displays correctly. The data itself is not damaged — you only need to change how it's read, not the original file.
: Yes. Upload the file to a large-CSV extraction tool, switch to column-extraction mode, and select the columns you need (multiple columns can be selected at once) to download a result containing only those columns. If you need only rows matching a condition, use row-matching mode and specify the condition column and value. Choose a tool that processes everything inside your browser without uploading the file to a server, and you can safely handle customer and sales data too.