lubridate mdy_hms() in R: Parse US Datetime Strings

The mdy_hms() function in lubridate parses month-first datetime strings (like "01/15/2024 14:30:00" or "January 15, 2024 2:30:00 PM") into a POSIXct value. It auto-detects separators, defaults to UTC, and accepts an Olson tz argument for any time zone.

mdy_hms("01/15/2024 14:30:00")                              # US slash, 24h time
mdy_hms("January 15, 2024 2:30:00 PM")                      # 12h with AM/PM
mdy_hms("01-15-2024 14:30:00")                              # dash separator
mdy_hms(c("01/15/2024 09:00:00", "02/20/2024 17:45:10"))    # vector input
mdy_hms("01/15/2024 09:00:00", tz = "America/Chicago")      # explicit timezone
mdy_hm("01/15/2024 14:30")                                  # no seconds
mdy_hms(x, truncated = 3)                                   # allow missing H/M/S
parse_date_time(x, orders = c("mdYHMS","mdYHM"))            # multi-format fallback

What mdy_hms() does in one sentence

mdy_hms("01/15/2024 14:30:00") parses the string and returns a POSIXct date-time. The "m-d-y" prefix tells lubridate the ordering is month, day, year; the "h-m-s" suffix says the input also carries hour, minute, and second. Output is POSIXct (a single instant plus a time zone), defaulting to UTC.

Use mdy_hms() when the source data writes dates US-style. Spreadsheets, CRM exports, US healthcare records, and many government feeds default to month-first. Pair the parser with the data convention; do not reformat the strings first.

Syntax

mdy_hms(x, tz = "UTC", locale = ..., truncated = 0, quiet = FALSE). x is a character vector or factor.

The return value is POSIXct, the same class returned by Sys.time(). Every downstream lubridate accessor (hour(), minute(), floor_date()) and every base-R operation (difftime, comparison, sorting) works directly on the parsed result.

Five common patterns

1. Standard US slash datetime

The most common case for US data. Output prints ISO-style ("2024-01-15") regardless of the input ordering; the underlying instant is what matters.

2. 12-hour clock with AM/PM

lubridate recognizes "AM" and "PM" tokens and shifts the hour accordingly. Noon parses as 12:00, midnight as 00:00. US-sourced logs and emails frequently use this format; do not hand-convert to 24-hour first.

3. Vector input with mixed separators

mdy_hms() is vectorized. Mixed slashes, dashes, and text months parse correctly inside a single vector, so messy CSV columns rarely need pre-cleaning.

4. Explicit US time zone

tz = "America/Chicago" labels the input as Central time. with_tz() then converts the same instant to UTC for storage or comparison. List valid zones with OlsonNames(); never use ambiguous abbreviations like "CST".

5. Inside a dplyr pipeline

After parsing, hour(), wday(), floor_date(), and the rest of lubridate's accessors chain naturally inside mutate(). The raw string column stays in place so any failed row can still be inspected.

mdy_hms() vs ymd_hms() vs dmy_hms() vs parse_date_time()

Pick the parser whose name matches the ordering in the source.

When two formats co-exist in one column, parse_date_time(x, orders = c("mdYHMS","ymdHMS")) tries each ordering in turn and keeps the first that succeeds. For everyday US data the dedicated mdy_hms() is faster and clearer.

Common pitfalls

Pitfall 1: ambiguous mdy versus dmy strings. mdy_hms("01/02/2024 09:00:00") returns January 2; dmy_hms("01/02/2024 09:00:00") returns February 1. The same string has two valid readings. Confirm the source convention by sampling 5 to 10 rows before choosing a parser.

Pitfall 2: silent UTC default. Without tz, the result is labeled UTC even if the data was captured locally. A 09:00 sales record from a Chicago store becomes "09:00 UTC" (which is 03:00 Central), six hours off. Pass tz = "America/Chicago" whenever the data is local.

Pitfall 3: silent NA on parse failure. A typo, an empty string, or a "N/A" placeholder returns NA with a single warning. Always run sum(is.na(parsed)) after parsing, then inspect x[is.na(parsed)] to see the offending rows. Date bugs almost always trace back to a silent NA at the parse step.

Try it yourself

Related lubridate functions

After mastering mdy_hms(), the most useful neighbors are:

• mdy_hm(), mdy_h(): month-first datetimes without seconds or minutes
• ymd_hms(), dmy_hms(): other date orderings with full time
• parse_date_time(): multi-order fallback for genuinely mixed inputs
• with_tz(), force_tz(): convert versus relabel a time zone
• hour(), minute(), second(), wday(): extract components from POSIXct
• floor_date(), ceiling_date(), round_date(): bucket datetimes to day, hour, minute
• now(), today(): current POSIXct or Date in the session zone

The official function reference lives at lubridate.tidyverse.org.

FAQ

How do I parse "01/15/2024 14:30:00" in R?

Use lubridate::mdy_hms("01/15/2024 14:30:00"). The function returns a POSIXct value defaulting to UTC. Pass tz = "America/Chicago" (or another Olson name) when the data is local time. No format = argument is needed; the parser auto-detects separators.

What is the difference between mdy and mdy_hms?

mdy() returns a Date object that stores year, month, and day only. mdy_hms() returns a POSIXct that stores a full instant including hour, minute, second, and time zone. Use mdy() for calendar dates and mdy_hms() for events with sub-day resolution like orders, logs, or appointments.

How do I handle AM/PM with mdy_hms?

mdy_hms() recognizes "AM" and "PM" tokens automatically. mdy_hms("01/15/2024 2:30:00 PM") parses to 14:30 in the result. Noon ("12:00 PM") stays at 12:00 and midnight ("12:00 AM") becomes 00:00. Do not hand-convert to 24-hour time before parsing.

Why does mdy_hms return NA for a valid-looking string?

The most common causes are an unsupported separator, a non-Olson timezone abbreviation in tz, or a day-first input misrouted to mdy_hms(). Run mdy_hms(x, quiet = FALSE) to see the warning, then sanity-check against OlsonNames() or fall back to parse_date_time(x, orders = c("mdYHMS","dmYHMS","ymdHMS")).

How do I convert mdy_hms output to a different time zone?

After parsing, use with_tz(parsed, "Europe/London") to convert the same instant to another zone (the moment in time does not change, only the displayed clock). Use force_tz(parsed, "Europe/London") to relabel without converting (the displayed clock does not change, the underlying instant shifts). The two functions answer different questions; pick deliberately.