Where to find free public datasets (CSV) for your projects

A good dataset makes a data project; a bad one sinks it. Whether you're practicing data analysis, writing a thesis or building a portfolio, here are the most reliable places to find free public data in CSV and similar formats, and a quick checklist to tell a usable dataset from a trap.

Search engines for data

Google Dataset Search

Dataset Search indexes datasets published across the web, from government portals to university repositories. Filter by format, license and date. It's the best starting point when you don't know who publishes the data you need.

dedigger

Researchers, students and agencies also share raw CSV and spreadsheet files on Google Drive. Our public datasets search filters for those formats. Add a topic, for example air quality or cricket. Files found this way often lack documentation, so check them carefully (see the checklist below).

Platforms with ready-to-use datasets

• Kaggle Datasets: thousands of datasets uploaded by the community, many with example notebooks. Great for practice; check the source of each one.
• UCI Machine Learning Repository: classic, well-documented datasets used in machine learning courses.
• Our World in Data: clean, sourced data on health, population, energy, economy and more, downloadable as CSV from every chart.
• Awesome Public Datasets: a long list on GitHub, organized by topic.

Official open data portals

• data.gov.in: India's open government data platform, with datasets from ministries and states.
• data.gov: United States federal open data.
• data.europa.eu: open data from the EU and its member states.
• World Bank Open Data: development indicators for every country, going back decades.

Most countries and many cities run similar portals; search for "open data" plus the name of the place.

Checklist before you use a dataset

1. Who made it? Prefer data with a named source you can check. If a file on Kaggle or Drive was copied from somewhere else, find the original.
2. What does each column mean? Look for a codebook, data dictionary or README. Without one, units and codes are guesses.
3. How old is it? Check the collection dates and whether it's still updated.
4. What's the license? Open licenses such as CC BY let you reuse data with credit. Some datasets are for non-commercial use only.
5. How clean is it? Open it and look at missing values, duplicates, odd outliers and inconsistent spellings before you trust any result.
6. Does it contain personal data? Data about identifiable people needs care, and sometimes permission, even if it's public. If you find personal data that looks leaked, don't use it.

Tools to open big CSV files

Google Sheets is fine for small files but has a limit of 10 million cells per spreadsheet and slows down long before that. For bigger files, use Python with pandas in Google Colab, which is free and runs in the browser. Our guide on learning Python for free covers where to start.

Turn practice into skills

Finding data is the first step; knowing what to do with it is what employers look for. If you want structured practice, DataCamp teaches data analysis through short exercises in the browser, and Coursera offers certificates such as Google's Data Analytics program. A portfolio of two or three projects on real public data, with a clear question and a clear answer, is worth more than a long list of courses.

Frequently asked questions

Can I use public datasets for commercial projects?

It depends on each dataset's license. Government data is often free to reuse; community uploads vary.

What's the best dataset for a beginner project?

Something you care about, small enough to open in a spreadsheet, with documentation. Our World in Data and Kaggle's popular datasets are good starting points.