CSV row counter: count the rows in any CSV file

How to count rows in a CSV file

1. 1Choose your file. Drag a CSV, TSV or TXT file onto the counter, or click “Choose a file”. Any size works, including files far too big for Excel.
2. 2Let it count. Your browser streams the file through a background thread and counts every line. A progress bar shows how far it has got, and you can cancel at any time.
3. 3Read the result. You get the number of lines, the delimiter, the number of columns in the first line, the file size and the time taken. Subtract one for the header line to get the number of data rows.

Lines, rows and records: what the number means

The counter reports physical lines: every line break starts a new line, and a final line without a trailing line break still counts. For most CSV files, lines and rows are the same thing, with a few exceptions worth knowing:

• The header line. Most CSV files start with a line of column names. If yours does, the number of data rows is the line count minus one: a file with 1,001 lines holds a header and 1,000 records.
• Line breaks inside quoted values. CSV allows a field to contain a line break if the field is wrapped in double quotes, for example a multi-line address or comment. Such a record spans two or more lines, so the line count is higher than the record count. A full CSV parser, like the one behind the upcoming MegaRows viewer, counts records instead.
• Blank lines count as lines too, which matters if an export ends with a few empty lines.

Here is a small example. The counter reports 4 lines, but the file holds a header and only 2 records, because Bob’s comment spans two lines:

id,name,comment
1,Alice,"Called on Monday"
2,Bob,"Line one
line two"

Windows (CRLF) and macOS or Linux (LF) line endings are both counted correctly. Very old Mac files that use a bare carriage return (CR) as the line break are rare today and are reported as a single line.

Built for very large files

The counter never loads the whole file into memory. Your browser hands it over in small chunks, a background thread (a Web Worker) counts the line breaks in each chunk and then lets it go, so memory use stays low however big the file is. Because the work happens off the main thread, the page stays responsive and you can cancel at any time.

In our tests on a recent laptop, a 10-million-line CSV of just over 400 MB is counted in about 0.6 seconds. Speed depends mostly on your disk and browser, so slower machines take longer, but there is no upload to wait for and no size limit to hit. If your file is too big for Excel, our guide on how to open a CSV file that’s too big for Excel walks through what to do next.

Other ways to count rows in a CSV

If you prefer the command line or need to count rows in a script, these work well:

macOS and Linux: wc -l

wc -l data.csv

Fast and built in. wc counts line-break characters, so it reports one less than the real number when the last line has no trailing line break.

Windows: PowerShell or Command Prompt

# PowerShell: streams the file, so it copes with big files (use the full path)
[Linq.Enumerable]::Count([IO.File]::ReadLines("C:\data\data.csv"))

# Command Prompt
find /c /v "" data.csv

(Get-Content data.csv).Count also works, but it reads the whole file into memory first, which gets slow beyond a few hundred megabytes.

Python

# Lines, like the counter above
with open("data.csv", "rb") as f:
    lines = sum(1 for _ in f)

# Records (handles quoted line breaks), header excluded
import pandas as pd
records = sum(len(chunk) for chunk in pd.read_csv("data.csv", chunksize=1_000_000))

Reading in chunks keeps pandas from loading a huge file into memory all at once. A plain len(pd.read_csv("data.csv")) is fine for smaller files.

Excel

Open the file and press Ctrl+End to jump to the last used row. This only works up to Excel’s limit: if the last row is 1,048,576, the file was almost certainly cut off and holds more rows than Excel shows.