Skip to content
AZ Tools

Julian Day Converter

Convert between a Gregorian date and a Julian Day (JD), Julian Day Number (JDN), Modified Julian Day (MJD), Reduced/Truncated JD, and Unix timestamp.

Couldn't read the Julian Day — must be a number.

Conversion uses the standard Meeus algorithm. Valid for dates from JD 0 (4713 BC) through the far future — leap-year rules and the 1582 Gregorian reform are both handled.

How to use

  1. Pick the direction. Date → JD takes a Gregorian date and time; JD → date takes a Julian Day number.
  2. Read the row that matches your downstream system — astronomy and satellite tools usually want MJD; observatory logs and ephemerides use full JD.
  3. Copy the value you need. The Unix timestamp helps when bridging to standard programming APIs.

About this tool

Show

Julian Day is a running count of days with nothing else in it: no months, no leap years, no calendar reform, no time zones. Day zero begins at noon on 1 January 4713 BC in the proleptic Julian calendar, and every instant since has exactly one number. That is why astronomy and historical chronology use it — subtracting two dates is subtraction, and a date from before the Gregorian reform needs no special handling because the count never noticed the reform happened.

The converter goes both ways and emits the variants you are likely to need. JD itself starts at noon, which is a deliberate choice from before electric light: an astronomer working through the night should not have the date change halfway through the observation. MJD subtracts 2400000.5 precisely to undo that, moving the day boundary back to midnight for satellite tracking and telemetry, with its epoch at 17 November 1858. Truncated JD and RJD are further shifts for systems that wanted smaller numbers. The Unix epoch sits at JD 2440587.5, and the J2000.0 standard epoch is JD 2451545.0 — noon on 1 January 2000, not midnight, for the same reason.

The matching Julian-calendar date is given alongside the Gregorian one because the two disagree, and the size of the disagreement changes over time. A date written by a contemporary of Kepler is not the date a modern calendar assigns to that day, and the gap was ten days when the reform began and is thirteen now. If you are converting a historical record, the calendar it was written in matters more than the arithmetic.

One caveat if precision matters. A Julian Date is only as well-defined as the time scale behind it, and the common scales differ by a fixed but non-zero amount: Terrestrial Time currently runs about 69.2 seconds ahead of UTC, because TAI is 37 seconds ahead of UTC and TT is 32.184 seconds ahead of TAI. Ephemerides are published in TT while your clock reads UTC, so a JD copied from one context into the other is a minute-scale error — irrelevant for a date, fatal for an occultation timing.

Frequently asked questions

Why does a Julian Day include a .5?

JD epochs start at noon UT, not midnight. So 2026-01-01 00:00 UT is JD 2461041.5. The half is what aligns 'days' to start when astronomers' observations begin.

Julian Day vs Julian calendar?

Totally different things despite the name. Julian Day is a continuous day count for time calculations. Julian calendar is the pre-1582 calendar (currently 13 days behind Gregorian). This tool shows both because the second is what historical astronomers used to record observations.

What is the point of counting days instead of using dates?

It removes calendars from the arithmetic. The interval between two events is a subtraction, with no months of different lengths, no leap rules and no calendar reform in the middle — which is why astronomy, and anything working with observations spanning centuries, keeps time this way. A date is for people; a day count is for calculation.

Which variant should I record?

Whichever the other side expects, and say which one you used. The plain Julian Day starts at noon and carries a large number; Modified Julian Day starts at midnight and subtracts a fixed offset to keep the number small — convenient, and the usual choice for satellite and geodesy data. The failure mode is silent: mixing the two shifts everything by 2400000.5, and a half-day offset is easy to mistake for a time-zone error.

Related tools