CRA Simplified Logbook: How the Base-Year Rule Works

Keeping a full vehicle logbook for twelve months is the single most abandoned habit in Canadian self-employment. The CRA knows this, which is why it publishes a way out: keep one complete year, then keep three months a year after that and project the rest.

The way out is real. But there is a test attached to it, and the test is widely repeated with the wrong reference point. Get the reference point wrong and you will believe a sample logbook supports your year when it does not, which is a problem you only discover during a review, years later, when the driving it was supposed to document is long gone.

This page states the rule the way the CRA states it, and works the arithmetic twice so you can check your own year against it.

What the simplified logbook actually is

Two steps, in order.

Step one, the base year. You keep a full logbook for one complete twelve-month period. Every business trip gets a date, a destination, a purpose and a distance, and the odometer is read at the start and the end of the fiscal period. That year gives you an annual business-use percentage, and it also gives you a business-use percentage for each quarter inside it. Both matter later. What a full logbook has to contain, field by field, is its own guide.

Step two, the sample years. In a later year you keep a logbook for a three-month sample period instead of the whole year. You do not use the sample percentage as your claim. You use it to compute a figure the CRA calls the calculated annual business use, and that figure is your claim, if it passes the test.

The base year does not stay valid forever by default. The CRA's condition is that you have to be able to show the use of the vehicle in the base year remains representative of its normal use. A business that changed shape (moved cities, dropped its delivery work, hired a driver) no longer has a representative base year, whatever the arithmetic says.

The formula

(Sample year period % ÷ Base year period %) × Base year annual % = Calculated annual business use

Read it as a scaling factor. The first bracket asks how this year's three months compare with the same three months in the base year, which is what strips seasonality out of the comparison. That ratio is then applied to the base year's annual figure.

Two things follow immediately, and both are where the rule gets misread.

1. The sample percentage is never your claim. It is one of three inputs.
2. The comparison that decides whether the result is accepted is against the base year's annual business use. Not the sample. Not the base year's quarter.

Worked example one: the CRA's own, which passes

This is the example on the CRA's motor vehicle records page, reproduced so you can see the shape of a year that works.

A full twelve-month logbook shows business use by quarter of 52%, 46%, 39% and 67%, and an annual business use of 49%. That is the base year.

In a later year the driver keeps a logbook for April, May and June only, and it shows 51% business use. In the base year, those same three months were 46%.

Input	Value
Sample period (Apr to Jun, later year)	51%
Same period in the base year (Apr to Jun)	46%
Base year annual	49%
Calculated annual business use	(51 ÷ 46) × 49 = 54%

Now the test. Is 54% within ten points of the base year's annual 49%? The band is 39% to 59%. It is. The CRA accepts 54% as the business use for that later year, absent contradictory evidence.

Notice what the arithmetic did. A slightly busier spring than the base year's spring (51 against 46) scaled the whole year up from 49% to 54%. The driver claims more than the base year, not the same, and the sample is what earned it.

Worked example two: the same machinery, failing

Change three numbers and the answer changes character.

Input	Value
Sample period (Apr to Jun, later year)	60%
Same period in the base year (Apr to Jun)	45%
Base year annual	40%
Calculated annual business use	(60 ÷ 45) × 40 = 53.3%

The test: is 53.3% within ten points of the base year's annual 40%? The band is 30% to 50%. 53.3% is outside it. The base year is not an appropriate indicator of annual usage in that later year, so the projection is not available.

What that actually costs you is specific, and worth reading slowly. It does not mean the sample was wrong or dishonest. It means the sample logbook is now reliable only for the three months it covers. For the other nine months you need an actual record of travel, or alternative records, neither of which you kept, because the entire point of the sample was to avoid keeping them. And the CRA's own guidance is that in these circumstances you should consider establishing a new base year by keeping a logbook for a fresh twelve-month period, which you cannot do retroactively either.

That is why the reference point matters. It is not a rounding argument. It decides whether nine months of your year has support.

The wrong denominator, and why it waves that year through

The version of this rule that circulates on tracker and bookkeeping guide pages tests those ten points against the sample period instead of against the base year's annual figure.

Run example two through it. The calculated figure is 53.3% and the sample was 60%. They are 6.7 points apart, comfortably inside ten. Under the wrong version the year passes and you file a 53.3% claim believing it is supported by a three-month log. Under the rule the CRA publishes, it failed, and the same claim is supported for one quarter and unsupported for three.

The two tests agree far more often than they disagree, which is exactly what makes the error durable. They come apart precisely when your driving pattern has moved away from the base year, which is the case the test exists to catch.

There is a second, gentler misreading that costs money rather than support: applying the base year's percentage straight to the later year instead of the calculated figure. On the CRA's own example that means claiming 49% when the arithmetic gives 54%. Five points of a full year of fuel, insurance, repairs and capital cost allowance, given back for no reason. The projection is allowed to move up as well as down.

Choosing the sample period

The formula compares your three months against the same three months of the base year, so picking your busiest quarter does not, by itself, inflate the result. A busy April in the sample year is measured against a busy April in the base year, and the ratio comes out near one.

Where it does bite is the ten-point test. A quarter that behaves unusually, in either direction, scales the whole year by that unusual ratio, and a big enough scaling pushes the result outside the band and costs you the projection entirely. A quarter that looks like your ordinary trading is the one most likely to survive.

Also worth saying plainly: the sample is three consecutive months of a complete log, not three months of remembering the good trips. Every business trip in the window, with all four fields.

How long you keep the base year

Ordinary records are kept six years from the end of the tax year they relate to. The base-year logbook is different, and the difference is easy to miss.

The full twelve-month logbook must be kept six years from the end of the tax year for which it was last used to establish business use. If you established a base year and then leaned on it for four more years of sample logbooks, the six years start counting at the end of that fifth year. In real time that is a decade of keeping one notebook, or one export, safe.

Practical consequence: the base year is the most valuable single document in your vehicle file. It is worth more than any individual year's log, because every later year's claim hangs off it.

Who this rule is for

The simplified logbook is about supporting a business-use percentage, so it reaches everybody whose claim runs on one.

• Self-employed and partners. The percentage goes onto your motor vehicle expenses on the T2125, applied to actual costs. If you want to see the arithmetic on your own numbers, and the difference between that and the per-kilometre rate people search for, the mileage calculator does both side by side.
• Employees with a signed T2200. Same percentage, applied to actual costs, on a T777. The per-kilometre rate is not available to you either.
• Incorporated owners paid a per-kilometre allowance. Your corporation is not applying a percentage, it is paying you per business kilometre, so it needs the kilometres themselves rather than a projection. The simplified logbook does not fit that case cleanly. See paying yourself a per-kilometre allowance for what the corporation actually has to hold.

FAQ

Is the ten percent test measured against the sample or the base year? Against the base year's annual business use. The CRA's example makes it explicit: a base year with 49% annual business use accepts a calculated annual figure anywhere from 39% to 59%, which is ten percentage points either side of 49%. The sample period's own percentage is an input to the formula, never the thing the result is compared to.

Is it ten percent or ten percentage points? Percentage points. The CRA writes "within 10%", and then works an example that resolves it: 49% annual gives a band of 39% to 59%. Ten percent of 49 would have given 44.1% to 53.9%, and the CRA does not use that.

What happens if the calculated figure falls outside the band? The sample logbook is reliable only for the three months it covers. The rest of the year has to be supported by an actual record of travel or alternative records, and you should consider starting a new base year with a fresh twelve-month logbook. This is the outcome you cannot fix after the fact, which is the argument for keeping a full log by default.

Can I use a sample logbook in my first year? No. The simplified method requires one complete twelve-month logbook first, to establish the base year. There is nothing to project from until that exists.

How long do I keep the base-year logbook? Six years from the end of the tax year for which it was last used to establish business use, which is later, often much later, than six years from the year it was kept. Ordinary records are six years from the end of the year they relate to.

Does a GPS app satisfy this? It satisfies the distance and the date. The CRA asks for the destination and the purpose of each business trip as well, plus the odometer at the start and end of the fiscal period, and a separate record per vehicle. An automatic trip capture with an empty purpose field is half a logbook.

Sources

1. CRA, Motor vehicle records, the full logbook requirements, the simplified logbook, the projection formula, the worked example, and the retention rule for the base year.
2. CRA Guide T4002, Self-employed Business, Professional, Commission, Farming, and Fishing Income, motor vehicle expenses for the self-employed.
3. Interpretation Bulletin IT-521R, Motor Vehicle Expenses Claimed by Self-Employed Individuals, archived, cited by the CRA's records page as a supporting publication.