Importing a CSV file into a database should be simple. In practice you spend time writing INSERT statements by hand, guessing column types, and wrestling with dialect differences between MySQL, PostgreSQL, and SQLite. This guide explains every step of CSV-to-SQL conversion so you can do it correctly — or skip the manual work entirely with our CSV to SQL converter.

What CSV to SQL Conversion Produces

A CSV file has two logical parts: a header row that names the columns, and data rows that hold values. Converting to SQL means generating:

  1. A CREATE TABLE statement that declares column names and types.
  2. One or more INSERT INTO statements that load the data.

Given this CSV:

id,name,email,age,active
1,Alice,[email protected],30,true
2,Bob,[email protected],25,false

The converter produces:

CREATE TABLE data (
  id INTEGER,
  name TEXT,
  email TEXT,
  age INTEGER,
  active BOOLEAN
);

INSERT INTO data (id, name, email, age, active) VALUES
  (1, 'Alice', '[email protected]', 30, TRUE),
  (2, 'Bob', '[email protected]', 25, FALSE);

How Header Detection Works

The first row of a CSV file is assumed to be a header. Headers become column names in the SQL schema. A good converter sanitizes these names:

  • Replaces spaces and special characters with underscores (order date → order_date)
  • Lowercases the names for consistency
  • Avoids reserved SQL keywords as column names (renames order to order_col, for example)

If your CSV has no header row, the converter should offer numeric column names (col1, col2, …) or let you enter column names manually.

Data Type Detection

Automatic type inference scans the values in each column and picks the most appropriate SQL type:

Detected patternSQL type
All values are whole numbersINTEGER
Values contain decimalsFLOAT / NUMERIC
Values are true/false (case-insensitive)BOOLEAN
Values match ISO 8601 date (2024-01-15)DATE
Values match ISO 8601 datetimeTIMESTAMP
Mixed or unparseableTEXT / VARCHAR

Type inference is best-effort. It scans every row, so a column is only typed as INTEGER if every non-empty cell is an integer. A single non-numeric value causes the column to fall back to TEXT.

Dialect Differences

MySQL

MySQL uses INT instead of INTEGER, requires backtick quoting for reserved-word identifiers, and uses TINYINT(1) for booleans in older versions.

CREATE TABLE `users` (
  `id` INT,
  `name` VARCHAR(255),
  `active` TINYINT(1)
);

PostgreSQL

PostgreSQL is strict about types. Use SERIAL or BIGSERIAL for auto-increment primary keys. Boolean literals are TRUE/FALSE. Text columns default to TEXT with no length limit.

CREATE TABLE users (
  id SERIAL PRIMARY KEY,
  name TEXT,
  active BOOLEAN
);

SQLite

SQLite uses dynamic typing. It accepts INTEGER, REAL, TEXT, BLOB, and NUMERIC. There is no native boolean — use INTEGER with values 0 and 1. SQLite also ignores column width declarations, so VARCHAR(255) and TEXT behave identically.

CREATE TABLE users (
  id INTEGER PRIMARY KEY AUTOINCREMENT,
  name TEXT,
  active INTEGER
);

Handling Common CSV Problems

Quoted fields — CSV values wrapped in double quotes can contain commas and newlines. A correct parser handles "Smith, John" as a single field, not two.

Missing values — Empty cells become NULL in SQL. The converter should emit NULL explicitly rather than an empty string.

Large files — Batch INSERTs perform better than one statement per row. The converter should group rows:

INSERT INTO data (id, name) VALUES
  (1, 'Alice'),
  (2, 'Bob'),
  (3, 'Carol');

Encoding — CSVs from Excel often use Windows-1252 encoding. If you see garbled characters, save the file as UTF-8 before converting.

Numeric strings — Zip codes, phone numbers, and product codes look like integers but must stay as text. A converter that detects leading zeros (07700) or phone format (+44-7911-123456) should leave these as TEXT.

Using the zerotool.dev CSV to SQL Converter

  1. Go to zerotool.dev/tools/csv-to-sql.
  2. Paste your CSV or upload a file.
  3. Choose your target dialect: MySQL, PostgreSQL, or SQLite.
  4. Review the detected column types — you can override them before generating.
  5. Copy the CREATE TABLE and INSERT INTO statements and run them in your database client.

No data is sent to any server. The conversion runs entirely in your browser.

When to Use CREATE TABLE vs INSERT Only

If the table already exists in your database, skip the CREATE TABLE statement and run only the INSERT INTO block. Most converters let you toggle this. If you are building a new table from scratch, generate both.

For large-scale imports, consider your database’s native bulk-load utilities instead:

  • MySQL: LOAD DATA INFILE
  • PostgreSQL: COPY FROM
  • SQLite: .import command in the CLI

These are faster than INSERT statements for millions of rows. For hundreds to thousands of rows, generated SQL works fine.

Try the CSV to SQL Converter →