CSV Schedule to Calendar: Bulk Add Events Without Losing Rows

Contents
If you've ever retyped dozens of events from an Excel or Google Sheets schedule into a calendar app one by one, you know how tedious it gets. Here's the short answer: a CSV table with many events can move into a calendar in one shot if you put one VEVENT block per row side by side inside a single .ics file.
The catch is that if column names or date formats don't match the expected rules during conversion, those rows drop out silently, with no message on screen. This post walks through the rules that decide which rows get recognized and which ones get skipped, based on how the converter actually behaves.
Summary ① Many events = many
VEVENTblocks in one.ics. Each block needs a differentUIDso it's treated as a separate event. ② Only two columns are required: title and start. Only the listed aliases are recognized. If neither column is present, the whole file fails to convert. ③ Dates are recognized only in year-month-day order with-,/, or.separators. If you leave the end column blank, it's calculated from the start value.
The hassle of copying events over one at a time
To get a schedule that's already in a table — a workshop agenda, a class timetable, a list of branch visits — into a calendar app, you usually have to type the title, date, time, and location for each event separately. Once you're at ten or twenty rows, that data entry alone eats a lot of time.
The iCalendar standard (RFC 5545) allows multiple BEGIN:VEVENT blocks inside a single BEGIN:VCALENDAR. So if you turn each row of your table into one VEVENT block and string them together, you can in theory build a single calendar file holding many events.
How to pack many events into one .ics by hand
The method itself is simple. Open with BEGIN:VCALENDAR, add a BEGIN:VEVENT through END:VEVENT block for each row of the table, then close with END:VCALENDAR. Fill each block with SUMMARY (title) and DTSTART/DTEND (start and end), and add LOCATION or DESCRIPTION if you need them.
One condition is easy to miss here: each VEVENT needs its own unique UID. When you build the blocks by copying one and changing the values, it's common to copy the UID along with it. Many calendar apps treat the same UID as an update to the same event, so some events may look like they disappeared.
With three or four rows this isn't hard. Past ten, generating a fresh UID for each one and normalizing the date formats is no different from retyping the whole table by hand.
Check it with our free tool. It handles the whole job — creating a new UID for every row and turning each one into a VEVENT block — so all you do is paste in your original CSV.
Free · no sign-up · conversion happens in your browser only, nothing is sent to a server

Common sticking points
| Situation | Cause |
|---|---|
| The whole file fails, with only an error message | The header has no column meaning "title" (title, subject, summary, event) and no column meaning "start" (start, start date, and so on) |
| Only certain rows silently disappear | The title cell in that row is empty, or the date in the start cell couldn't be recognized. The number of skipped rows shows on screen as "N rows skipped" |
| You entered a clear date but it isn't recognized | Only year-month-day order with -, /, or . separators is recognized. Something like "July 15, 2026" with letters mixed in, or month-day-year order, won't be recognized |
| You left the end column blank | Timed events end one hour after the start; all-day events end the following day. That's filled in automatically |
| An all-day event's end date looks one day longer | The date you put in the end column is treated as the event's last day, so the next day is calculated as DTEND (in iCalendar, DTEND is an exclusive end and isn't included) |
If your header uses a name that isn't in the alias list (say, "item name" or "date"), that column won't be recognized. Change just the first row's column names to expressions that include title and start, and everything else converts again as-is.
Conversion runs entirely inside your browser, from CSV parsing to .ics generation, and the event details you enter are never sent to or stored on a server. Times are saved in the device's local time (floating time) unless you specify otherwise.
Recap
- To move many events over, turn each row of your table into one
VEVENTblock and give every block a differentUID. - Only two columns are required, title and start, and only the listed aliases are recognized (for example, title or subject, start or start date).
- Dates are recognized only in year-month-day order with
-,/, or.separators. No time means an all-day event. - If you leave the end column blank, timed events get one hour after the start, and all-day events get the next day.
- Conversion happens only in your browser, and the schedule data you enter is never sent to a server.
Free · no sign-up · conversion happens in your browser only, nothing is sent to a server
Frequently asked questions
Do the column names have to be in English, like title and start?
No. For the title column you can use title, subject, summary, or event instead of title, and for the start column you can use start, start date, or start time instead of start. Case and leading or trailing spaces don't matter. A name that isn't on this list won't be recognized.
If I leave the end column blank, how is the time filled in?
For events whose start value includes a time, a blank end column gets the start time plus one hour automatically. For all-day events with a date but no time, the end becomes the next day, which makes it a one-day event.
I put thirty rows in my table, but the result says only twenty. Why?
Rows with an empty title or start value, or with a date format that couldn't be recognized, are dropped from the conversion, and the count shows on screen as "N rows skipped." Completely empty rows (every cell blank) aren't included in that count, so it may not match your original row count exactly. To find the missing rows, first check that the title and start cells are filled in and that the dates are in year-month-day format.
Is my converted schedule data sent to a server?
No. CSV parsing, .ics generation, and the reverse conversion all run in JavaScript inside your browser, and what you enter is never sent to or stored on a server. Even when you load a file, it's only read in your browser's memory.
This article was adapted from our Korean original. Read the Korean original