Plenty of text message disputes are not about what was said but about when. A message sent just before midnight on the night of a missed exchange, a “Read” receipt the other side says proves notice, a screenshot that says “Yesterday” with no date at all: timestamps look like the most objective part of a message exhibit, and they are among the easiest parts to get wrong.
This guide explains which timestamps a message carries, where each one comes from, how exports store them and how to normalize them into a single stated time zone that holds up under cross-examination. For the layout of the exhibit itself, see our guide to text message exhibit formatting.
Which timestamps does a text message actually have?
Usually more than one, and they can disagree. Depending on the platform and the settings on both phones, a single message can carry several recorded times:
| Timestamp | What it records | Where it comes from | Watch for |
|---|---|---|---|
| Sent | When the message was recorded as sent | The sender’s device clock, and for iMessage or RCS the service | A sender clock set manually or wrong |
| Received or stored | When the recipient’s phone logged the message | The recipient’s device clock; SMS also carries a network service-center time | A phone that was off or out of coverage receives late |
| Delivered | Confirmation the message reached the recipient’s device | iMessage and RCS delivery status; optional SMS delivery reports | Often shown only under the latest message in a thread |
| Read | When the recipient’s app marked the message read | The recipient’s app, only with read receipts on | Optional, can be off per conversation |
| Edited or unsent | When a message was changed or withdrawn | Newer versions of iMessage and some chat apps | Edit history may show on the device but not in every export |
| Reaction | When a tapback, emoji reaction or reply was added | A separate event with its own time | May arrive as a separate text line on another platform |
Exports rarely include every one of these, and screenshots usually show only one. The first job is knowing which time a given exhibit actually displays, and labeling it.
Where do the times in an export come from?
Most export tools read a message database from the phone or a backup and convert stored numbers into readable dates. The raw formats differ by platform:
- iPhone Messages. The Messages database counts time from January 1, 2001, in UTC, which is Apple’s reference date. Older versions store seconds and newer versions store nanoseconds, with separate fields for sent, delivered and read times where the service recorded them. See exporting iMessages for litigation.
- Android SMS and MMS. The system message store generally keeps milliseconds since January 1, 1970, in UTC. A received SMS often has two values: when the phone stored it and a sent time supplied by the network. Backup apps such as SMS Backup & Restore write both into their XML, plus a readable date in the phone’s zone. See exporting Android text messages.
- WhatsApp and similar apps. Chat exports are plain text, with dates in the exporting phone’s display format and zone and no offset stated.
- Carrier records. Call and text detail records come from the carrier’s own systems, often in UTC or in a single fixed zone the carrier chooses. The records key or the custodian should say which; see subpoenaing phone records.
Whenever a tool converts these values into a readable date, it applies a time zone, usually the zone of the computer running the tool or of the phone that made the export. Record which one, along with the tool and version, in the collection log.
Why don’t the times on a screenshot match the export?
Because a screenshot shows how the phone displayed the time at the moment of capture, not the stored value. The usual causes:
- Relative labels. Messages apps show “Today,” “Yesterday,” a weekday name or a date without the year, and print headers only after a gap. Every relative label is calculated from the day the screenshot was taken.
- The current zone, not the original one. Phones generally display stored times in the zone the device is set to when someone looks. A client who traveled, moved or changed settings before taking screenshots can show every message shifted by an hour or more.
- Date and clock formats. Day-first and month-first date orders, and 12- versus 24-hour clocks, can make the same digits mean two different days.
- Hidden per-message times. On iPhone, individual message times appear only when the view is swiped; many Android apps show them only after a tap. Most screenshots capture only the header.
- Status lines. “Delivered” or “Read” with a time usually belongs to the latest message in the thread, not to every message above it.
How do time zones and daylight saving time distort a timeline?
They move messages across midnight, and in custody matters across parenting days, which is where many arguments start. Check three patterns before any chronology goes to the attorney:
- Daylight saving transitions. In spring one local hour does not exist; in fall one hour repeats. Two messages stamped 1:30 a.m. on the fall change day could be an hour apart in either order. Only a UTC value or a stated offset settles it.
- Travel and relocation. If a party moved between zones during the period, pick one display zone for the exhibit and state it. When local time at the sender’s location matters to the facts, add it in a separate labeled column rather than mixing zones in one.
- Mixed-source chronologies. A chronology that combines a phone export in local time, carrier records in UTC and photo metadata with no offset can show events out of order. Convert every source before merging; see how to build a case chronology.
What does a read receipt show, and what doesn’t it?
A read receipt shows that the recipient’s messaging app marked a message as read at a particular time. Keep these limits in view:
- Receipts are optional. iPhone users can turn read receipts on or off for every conversation or for one at a time, and RCS chats on Android have similar settings. A missing receipt is not evidence that a message went unread.
- Marked read is not the same as read. Opening the thread, acting on a notification or opening the chat on a linked tablet or computer can mark messages read. A lock-screen preview can show the text without marking anything.
- SMS has no read status. When a message goes as ordinary SMS, shown as a green “Text Message” on iPhone, there is no read receipt at all. SMS delivery reports, where the sending phone requests them, show only that the network handed the message to the phone.
- One account, many devices. A message read on any device signed in to the same account may show as read everywhere.
How much a receipt matters is an argument for counsel. Staff should report what the record shows in the record’s own words, such as “Read 9:14 PM,” and avoid restating it as “she saw it.”
How do you normalize timestamps for an exhibit?
Keep the raw values, convert once and document every step so anyone can repeat the work:
- Preserve the raw export. Record a hash of each original file on arrival and work from copies; see hash values in litigation.
- Identify each time field. For every source, note which value you are using, such as sent, stored or read, and what the field represents.
- Convert to UTC first. Translate every raw value to UTC before choosing a display zone, so daylight saving and travel problems become visible instead of hidden.
- Choose one display zone. Usually the zone where the events happened or where the case is pending, applied to every message on the exhibit.
- Flag anomalies. Replies that predate the message they answer, messages out of sequence or a cluster with an impossible gap suggest a device clock problem or a merge error. Investigate before anything is produced.
- Cross-check independent sources. Call logs, carrier records, photo metadata, emails and calendar entries can confirm or contradict a disputed time.
- State the method on the exhibit. Name the zone, the field shown and the source, in a header or note.
Message Extractor merges a client’s exports into one chronological thread model, stamps every message with sender, timestamp and channel, and lists each source file with its SHA-256 value in the manifest, so the time basis for every page can be traced back to a specific export.
Template: time basis note for a message exhibit
All times on this exhibit are shown in [time zone] (UTC[offset]; UTC[offset] after [date] when daylight saving time [began/ended]). Times shown are the [sent/stored] times recorded in [source export, for example an iPhone backup export made with (tool and version) on (date)]. Read times appear only where the recipient’s device recorded a read receipt. Original values are preserved in UTC in the source file listed in the manifest, [file name], SHA-256 [value].
Which timestamp mistakes cause the most trouble?
Most chronology disputes trace back to a short list of avoidable errors:
| Mistake | Consequence | Fix |
|---|---|---|
| No time zone stated | Messages argued onto the wrong day | Name the zone and offset on every exhibit |
| Screenshot dates taken at face value | “Yesterday” decoded from the wrong day | Replace with an export; use capture metadata only to decode |
| Mixing local time and UTC sources | Events out of order in the chronology | Convert every source to UTC before merging |
| Thread-level “Read” applied to every message | A claim the record does not support | Attribute receipts only to the message they belong to |
| Ignoring daylight saving changes | Two messages in the wrong order | Keep UTC values and offsets for transition days |
For collection steps, see producing text messages in discovery. Decisions about what a timestamp means for the case, which zone to argue from and whether to retain a forensic examiner for a disputed device clock belong to the supervising attorney, and local practice on chronology exhibits should be confirmed with counsel licensed where the matter is pending.
