Does the CRA Accept a Mileage App or Only a Paper Logbook?
The CRA sets what a vehicle log must say, not what it is on. The content rule, the electronic record rules, and how paper, spreadsheets and apps each fail.
People ask whether the CRA accepts a mileage app as if the answer were a certification: some apps approved, some not, paper as the safe default. There is no such list. The CRA does not approve logbook formats, apps or notebooks. It says what a vehicle record has to contain, and it has separate, general rules for any record kept electronically.
So the CRA accepts a mileage app, a spreadsheet and a paper notebook on the same terms, and rejects each for the same reason: a log that does not say what it has to say, or cannot be produced when asked. What differs between formats is the way each one tends to fail.
The CRA prescribes content, not format
The CRA's motor vehicle records page describes the best evidence of business use as "an accurate logbook of business travel maintained for the entire year". For each business trip it asks for four things:
| Field | What it has to hold |
|---|---|
| Date | The day of the trip |
| Destination | Where you went |
| Purpose | Why the trip was business |
| Kilometres | The distance driven |
On top of the trips, it asks for the odometer reading of each vehicle at the start and the end of the fiscal period, the dates and odometer readings when you buy, sell or trade a vehicle during the period, and a separate record for each vehicle if you use more than one. What each of those means in practice has its own guide.
Nothing on that list names a medium. Subsection 230(1) of the Income Tax Act requires records "in such form and containing such information as will enable the taxes payable under this Act" to be determined, and the CRA accepts records kept on paper, on paper later converted to a readable electronic format, or in an accessible and readable electronic format.
Whatever the format, its job is one number: business kilometres over total kilometres. As a sole proprietor or partner, your T2125 claim is your actual vehicle costs multiplied by that percentage. The CRA's bulletin IT-521R says a claim calculated on a cents-per-kilometre basis is not acceptable; the per-kilometre rate is a ceiling on an allowance a payer pays, and who can use it is a short list.
The rules for a log kept electronically
An electronic log has a few obligations paper does not, and they are the part people skip when they assume an app takes care of everything.
Readable for the whole retention period. Subsection 230(4.1) of the Income Tax Act says a person who keeps required records electronically "shall retain them in an electronically readable format" for the retention period. Printing the log does not discharge that: the CRA says records originally produced in electronic format have to be kept in an electronic readable format "even if you have paper printouts".
Six years, and sometimes much longer. Subsection 230(4) sets the ordinary retention period at six years from the end of the last taxation year the records relate to. A base-year logbook used to support later sample years is kept six years from the end of the last year it was used for, which can run to a decade.
In a format the CRA can open. Information Circular IC05-1R1, Electronic Record Keeping, says the copy you provide must be "in a non-proprietary, commonly used data interchange format that is compatible with CRA software". A log that only exists inside an app, viewable on its screens but not exportable, is the weak case. A spreadsheet or CSV export is the strong one.
Backed up. The CRA's record-keeping page asks you to "always store proper backup copies, preferably at a site in Canada other than your business location". IC05-1R1 adds that backups should be tested periodically to confirm they can actually be restored.
Your responsibility, even when a company holds the data. IC05-1R1 says contracting record keeping out to a third party does not relieve you of any readability, retention or access responsibility, including through "bankruptcy or migration to/from a third party", and calls it good practice to get your own readable copy from the provider. The CRA also says records "kept outside of Canada and accessed electronically from Canada are not considered to be records kept in Canada". In plain terms: if an app holds your log, export it at least once a year and keep the file yourself, in Canada.
Contemporaneous versus reconstructed
The CRA asks for a log "maintained for the entire year". The Tax Court has spoken of a logbook "created at the time". That ranking matters more than the format. Two informal procedure decisions illustrate it; under section 18.28 of the Tax Court of Canada Act they are not precedents, but they show how the Court weighs a log.
- In Dore v. The Queen, 2004 TCC 638, the taxpayers had stopped keeping mileage logs, kept partial logs for part of one year, and then estimated their business use. Justice Bowie called the result "simply a guesstimate based on partial information" and said that estimates made at year end "are only that, estimates. Generally, they tend to be generous to the estimator." The auditor had extrapolated from the partial logs that did exist, and the Court upheld her figure.
- In Glawdecki v. The Queen, 2010 TCC 650, the taxpayer produced spreadsheets after the audit, built from his work time sheets and distances he worked out with GPS tracking software. He acknowledged he had not used a GPS device at the time and had not created the logs when he drove. The Crown called them vague about where he travelled and why. Justice Lamarre agreed that "in the absence of a logbook created at the time the expenses were incurred, the appellant has an evidentiary hurdle that is not easily overcome."
GPS-derived distances did not help in the second case, because they were produced afterwards and carried no destination or purpose. The same measurement taken while driving, with the purpose written against it, is the log the Court was asking for. If you have already lost a year, the reconstruction method is its own guide.
What paper tends to get wrong
Paper is accepted, and a good paper log is excellent evidence. Its failures are human, not technical.
Gaps. A notebook in the glovebox is filled in when you remember it is there. Two weeks in August go blank. A gap does not prove you did not drive for business, but it is visible, and every trip inside it is unsupported.
Written up later. Many paper logs are written at the kitchen table, a week or a quarter at a time, from memory. That leaves the tells a reviewer notices: the same pen on every page, round distances, the same trip every weekday with no variation, purposes repeated word for word, and distances guessed rather than mapped. A log written up later is a reconstruction presented as a record kept at the time.
Survival. One notebook, kept six years and possibly ten. The CRA says that generally you must keep paper originals, and that images replace them only when made to the national imaging standard it names. Scanning each page is a sensible backup, but keep the notebook too.
What apps tend to get wrong
An app that records drives automatically solves the paper problems: no gaps from forgetting, measured distances, a timestamp on every trip. Its own failures are easy to miss because the log looks full.
A captured trip with an empty purpose is half a log. GPS gives you date, start and end points, and distance. It cannot give you purpose. If the trip was classified as business with a tap, and nothing records why, a reviewer is looking at a drive to an address and your word that it was work. The purpose field is also the one that separates a business trip from the commute, which is where most overstated logs go wrong.
Sorting in bulk at year end. The trips were recorded at the time, but the judgement about each one is made months later. Sort each drive close to when you drove it.
The app's total is not your total. An app sees the drives your phone was in. It does not see a drive when the phone was at home, switched off, or with someone else driving your car. If you divide business kilometres by the app's total instead of the odometer's, the denominator is too small and the percentage comes out too high. That is the job the odometer readings the CRA asks for do. Read and keep them yourself, at both ends of the year.
Trips in the wrong car. Your phone rides in every vehicle you do. If you use more than one for the business, each needs its own record, and trips assigned to the wrong one corrupt both percentages.
An export with no audit trail. People export the year to a spreadsheet and then edit it: delete trips, retype a purpose, change a distance. The result has no record of what changed. IC05-1R1 is written for accounting systems rather than mileage logs, but it states the CRA's view of a reliable electronic record: a change to a recorded transaction should show who made it, when, the previous details, the new details, and why. Keep the original export beside any edited copy.
A worked year: what the empty purpose field costs
Take a sole proprietor with one vehicle and an app that recorded every drive. These are the inputs, stated so you can check the arithmetic:
- Odometer at the start of the fiscal period: 42,180 km. At the end: 66,480 km. Total driven: 24,300 km.
- Trips marked business in the app: 11,340 km.
- Of those, trips with a purpose recorded (client name, job, supplier): 8,910 km. The other 2,430 km were swiped business with no purpose written.
- Actual vehicle costs for the year (fuel, insurance, repairs, interest and capital cost allowance, all worked out on their own rules): $9,000.
| Business km | Business use | Vehicle claim | |
|---|---|---|---|
| Every trip marked business | 11,340 | 11,340 ÷ 24,300 = 46.7% | $9,000 × 46.7% = $4,200 |
| Only trips with a purpose | 8,910 | 8,910 ÷ 24,300 = 36.7% | $9,000 × 36.7% = $3,300 |
| Difference | 2,430 | 10 points | $900 |
(The exact fractions are 7/15 and 11/30, which is why the claims come out to round dollars.)
The 2,430 km are not automatically lost, and may well have been business. But they rest on nothing except the tap that sorted them, which makes them the easiest part of the claim to question and the hardest to defend: $900 of deduction resting on a blank field.
Now run the same year the other way. Suppose the app recorded only 21,600 km in total, because the car was sometimes driven without the phone in it, and you used that as the denominator: 11,340 ÷ 21,600 is 52.5%. The odometer says 46.7%. The app's own total overstates the percentage by almost six points, which is the kind of error that reads as an inflated claim even though every business trip in it was real. The odometer readings are what keep the percentage honest.
Paper, spreadsheet and app compared
| Paper notebook | Spreadsheet typed by hand | GPS app | |
|---|---|---|---|
| Accepted by the CRA | Yes | Yes, as an electronic record | Yes, as an electronic record |
| Date and distance | Written by hand; distance often guessed | Typed; distance from a map or a guess | Recorded automatically and measured |
| Destination | Written by hand | Typed | Recorded as a location; may need a name added |
| Purpose | Written by hand | Typed | Only if you add it |
| Typical failure | Gaps, and pages written up later from memory | Filled in at month end or year end | Trips sorted but no purpose; the app's total used instead of the odometer |
| Odometer anchors | You read and write them | You read and type them | You read and record them; the app cannot see the odometer |
| Evidence it was kept at the time | Weak on its own | Weak on its own (file dates can be changed) | Strong for the trip itself; timestamps come from the drive |
| Retention risk | Loss or damage of one notebook | Loss of one file | The app or the company going away |
| What protects it | Scan pages as a backup and keep the original | Backups in two places | Regular exports kept on your own storage |
What to look for in any method
Whatever you use, check it against this list. If you want a blank layout to start from, the mileage log template has one.
- All four fields, every business trip. Date, destination, purpose, kilometres. If your method does not make purpose easy to enter, it will be the field that goes missing.
- Captured at the time. Recorded as you drive, or on the same day. The CRA's "maintained for the entire year" and the Court's "created at the time" point the same way.
- Odometer at both ends of the fiscal period, and at any vehicle change, recorded by you from the dashboard. This is your denominator, and nothing else supplies it.
- One record per vehicle. Every trip tied to the car it was driven in.
- Exportable in a common format. A spreadsheet or CSV you can open without the app. IC05-1R1 asks for non-proprietary, commonly used formats.
- Kept, and backed up, for the whole retention period. Six years from the end of the last year the log relates to, longer for a base year. Keep a copy yourself, in Canada.
- Edits you can explain. If you change a trip later, keep the original or a note of what changed and why.
A method that meets all seven gives the CRA what it asks for, whatever it is written on. If you have settled on an app, the best mileage tracker apps in Canada compares six of them on the percentage, the odometer and what their free tiers keep.
Claimable Drive records drives in the background, you swipe each one business or personal, it works out the business-use percentage from those swipes and keeps the start and end odometer readings, and the logbook and its export are free.
FAQ
Does the CRA accept a mileage app? Yes, on the same terms as paper. The CRA does not approve or certify apps. It asks for the date, destination, purpose and kilometres of each business trip, the odometer readings at the start and end of the fiscal period, and a separate record per vehicle, and it accepts records kept in an accessible and readable electronic format. An app log that holds all of that, and that you can still produce years later, is acceptable.
Is GPS mileage tracking CRA compliant on its own? Not on its own. GPS captures the date, the locations and the distance. It cannot capture the purpose, and it does not read your odometer. A GPS log with purposes added and odometer readings kept is a complete log. A GPS log without them is half of one.
Is a paper logbook safer than an app? No. A paper log written as you go is good evidence. One written up from memory at month end is a reconstruction dressed as a record. What matters is whether the log was kept at the time and holds every field.
Can I print my app log and delete the app? Printing is not enough on its own. Subsection 230(4.1) of the Income Tax Act requires records kept electronically to be retained in an electronically readable format for the retention period, and the CRA says that applies even if you have paper printouts. Export the log to a common file format and keep that file for the full period.
I edited my exported spreadsheet. Is that a problem? Correcting a genuine error is fine. The weakness is an edited file with no trace of what it said before, so keep the original export beside the corrected one.
Sources
- CRA, Motor vehicle records, the logbook as the best evidence, the four trip fields, the odometer requirement, vehicle changes, one record per vehicle, and retention of the base-year log.
- Income Tax Act, section 230, subsection 230(1) (form and content of records), 230(4) (six-year retention) and 230(4.1) (electronic records retained in an electronically readable format).
- CRA, Acceptable format, imaging paper documents and backing up electronic files, paper and electronic formats accepted, electronic records kept electronically even with printouts, keeping paper originals, and backups.
- CRA, Your responsibilities and the requirements associated with records the law requires you to keep, records must be reliable and complete, electronic records readable by CRA software, and backup copies preferably at a site in Canada.
- CRA, Where to keep your records, for how long and how to request the permission to destroy them early, records kept in Canada, records accessed from outside Canada, and the six-year period.
- CRA Information Circular IC05-1R1, Electronic Record Keeping, paragraph 7 (non-proprietary, commonly used format), paragraphs 17 and 19 to 22 (backups, third-party providers), and paragraph 40 (controls over changes to recorded transactions).
- Interpretation Bulletin IT-521R, Motor Vehicle Expenses Claimed by Self-Employed Individuals, archived, paragraph 8: a cents-per-kilometre claim is not acceptable, and the record of total and business distance.
- CRA Guide T4002, Self-employed Business, Professional, Commission, Farming, and Fishing Income, motor vehicle expenses on the T2125.
- Dore v. The Queen, 2004 TCC 638, paragraphs 20 and 21.
- Glawdecki v. The Queen, 2010 TCC 650, paragraphs 8, 13 to 15.
- Tax Court of Canada Act, section 18.28, informal procedure judgments are not precedents.
Claimable Drive keeps this logbook for you.
An iPhone logbook, free for thirty drives a month, that records without being opened, suggests which drives are business, and shows the percent a self-employed Canadian actually deducts.
Get the free logbook