What to Plant in the GTA Month by Month (Toronto Hardiness Zone 7a to Georgina 5b)

Toronto sits in Canadian plant hardiness zone 7a on the current map, the 1991-2020 update published by Natural Resources Canada. It is not 5b, and it is not 6a. Those are the numbers from the 1961-1990 map, still printed in a lot of gardening books. The rest of the GTA is not 7a either: the region spans 5b in Georgina to 7a on the lakeshore, and the frost dates that actually decide when you plant differ by 47 days across it. There is no single GTA planting date. There are two numbers for your own corner of it, and every date in the calendar below is derived from them.

Your zone is probably not the number you were told

Natural Resources Canada runs the lookup that settles this. Its Plant Hardiness Zone by Municipality tool takes one municipality at a time and returns four values: the zone under the 1961-1990 map, the zone under the current 1991-2020 map, and the underlying hardiness index for each. Queried town by town across the region, it gives this.

Two patterns run through it. Every municipality in the table that reached 7a (Toronto, Mississauga, Oakville, Burlington) sits on the Lake Ontario shoreline. And Georgina, at the top of the region on Lake Simcoe, is the only one of the twenty-three that has not changed zone at all since the 1961-1990 map.

The index column tells you how securely a place sits in its band. Natural Resources Canada calculates a suitability index from seven climate variables, then groups the result: "index values of 40 to 44 corresponds to zone 4a, index values of 45 to 49 corresponds to zone 4b, 50 to 54 to zone 5a and so on." Follow that pattern and 70 to 74 is zone 7a. Toronto's 71 is one point above the floor of its band, and Mississauga's 70 is exactly on it. Treat 7a as a recent and marginal designation, not a licence to plant like Niagara: a single point of index is the difference between 7a and back to 6b.

Canadian zones and USDA zones are not the same thing

This is where most gardening advice goes wrong, because most of it is American, and the two systems measure different things.

The USDA map, as Natural Resources Canada describes it, "averages the lowest annual minimum temperature over a standard time period" and classifies the result into "10-degree F (5.6 degree C) zones and 5-degree F (2.8 degree C) subzones." One variable. Canada's map uses the seven-variable index above, folding in the frost-free period, summer maximum temperatures, June-to-November rainfall, snow depth and the maximum wind gust in 30 years alongside winter cold.

Natural Resources Canada ran the comparison and its conclusion is blunt: there are "no simple conversions between the two systems." When the two maps are overlaid, Canadian hardiness zone 4a overlaps nine different USDA zones, ranging from 3a to 7a. So a Toronto gardener reading "hardy to zone 7" on an American seed rack has learned nothing reliable about their own garden, in either direction.

The two dates the whole calendar hangs on

Zone tells you what survives the winter. It does not tell you when to plant. That comes from the frost dates in Environment and Climate Change Canada's 1991-2020 Canadian Climate Normals, and the GTA's three long-running stations disagree sharply.

Downtown against Markham is 208 - 161 = 47 days of growing season between two stations about 20 km apart. A downtown raised bed and a Markham back garden are not running the same calendar, and any single date offered for the region as a whole is wrong for at least two of these three places.

Read the right-hand column before you lean on that gap, because the three records are not equivalent. Environment and Climate Change Canada codes each normal by how much data stands behind it: "B = At least 25 years, C = At least 20 years, D = At least 15 years." Pearson and Buttonville are coded C. The downtown figures are coded D, and that frost record ends in 2011, not 2020. So the 47-day spread compares a 1991-2011 downtown record against two 1991-2020 suburban ones. The agency's own note makes the point: "a composite station with 30 years of record between 1991 and 2020 with no missing years would be a more reliable normal than a composite station with 15 years of record and 2 missing years." Downtown genuinely is the warmest of the three. Some of the margin is the record, not the climate.

Averages are not deadlines, and the two tables are not comparable

Each station also publishes a probability distribution, and that is the part that decides whether you gamble. At Pearson, the probability that the last spring frost lands on or after a given date runs 50% for May 4, 25% for May 13 and 10% for May 19. Plant tomatoes out on the average date of April 27 and better than an even chance of frost is still ahead of you. Wait until May 19 and the risk is down to about one year in ten. Buttonville runs a few days behind at every step: 50% for May 5, 25% for May 11, 10% for May 18.

One trap comes with that table. The probabilities and the averages are computed over different periods at every station. Pearson's average last-frost date uses 1991-2020, 23 years of it; its probability rows use 1941-2020, 71 years: eight decades including much colder mid-century springs. Downtown the mismatch is wider still: an average from 1991-2011 against probability rows drawn from 1841-2020, 169 years. Buttonville's rest on 1987-2020, 25 years. So do not read "50% chance of frost on or after May 4" as contradicting "average last frost April 27." They are two records answering two questions, and the probability rows lean colder because they reach further back.

Turning two dates into planting dates

Everything below is arithmetic on those two numbers. Do it with your own station's dates rather than the region's.

Indoors, count backwards from your last frost. The seed packet gives the interval. The Toronto Master Gardeners put parsley at "about eight weeks before the last frost" ("in colder climates like ours"), lemon balm at six weeks, and French sorrel at three. Take parsley's eight weeks, which is 56 days, and subtract:

• Downtown: April 11 - 56 days = February 14
• Pearson: April 27 - 56 days = March 2
• Buttonville: May 3 - 56 days = March 8

Same seed, same instruction, twenty-two days apart across one region.

Outdoors, count forwards. The Toronto Master Gardeners' urban vegetable guide gives direct-seeding dates for Toronto in four bands: frost hardy (lettuce, peas, radishes, spinach) April 15-25; semi-frost hardy (beets, carrots) April 25-May 10; semi-frost tender (snap beans, tomato transplants) May 15-25; frost tender (transplanted cucumbers, peppers, squash) May 25-June 5. Those are Toronto dates. If your station's last frost is later than downtown's, shift the whole set later by the same margin: about two weeks at Pearson, about three at Buttonville.

At the end of the season, count backwards from the first frost. Subtract the days-to-maturity on the packet from your station's first-frost date to get the last useful sowing day. At Buttonville, October 13 - 60 days lands on August 14. At the downtown station, November 6 - 60 days gives September 7: twenty-four extra days of sowing window, from the same seed packet.

Two conditions travel with all of it. Seedlings raised indoors must be hardened off first. The Toronto Master Gardeners describe a process "over about ten days to two weeks," starting with "two or three hours a day" in a shaded, protected spot. And their guide is explicit that you can plant earlier than these dates "if you will be around to cover your plants if there is a cold snap, and if your garden is small enough that you can cover the plants."

The month-by-month calendar

Dates are keyed to Toronto Pearson, the middle of the three stations. Shift about two weeks earlier downtown and about a week later in Markham. Newmarket, Uxbridge and Georgina have no station of their own; their hardiness indices sit below Markham's, so treat them as later again rather than assuming Markham's dates.

Three autumn deadlines worth treating as real

Garlic. Ontario's garlic production factsheet says garlic here "is grown as a winter annual - planted in the fall and harvested the following summer," and that "it is critical that garlic not be planted too early or too late." Both errors are named. Too early or not deep enough, and "there is a risk that shoots will emerge above the soil surface and be prone to winter injury." Too late, and "cloves will not develop adequate root systems and will not be winter hardy." For mulch, the factsheet describes straw as the most commonly used, "applied 10-15 cm deep directly over the planted garlic rows."

Bulbs. The Toronto Master Gardeners' autumn guide says to "begin planting bulbs in September," with narcissus, allium and small bulbs "planted by the end of September or early October." Tulips are the exception: they "can be planted as long as the ground can still be worked before freeze-up."

Lawn and mulch. The same guide puts lawn re-seeding in early September, because "after September the lawn needs a slow growth period to harden off for winter", and if you miss the window, its advice is to "wait for spring" rather than seed late. It names the first week of November as the best time in Toronto to fertilise a lawn. And the mulch instruction carries a condition people routinely drop: mulch tender perennials after the ground has frozen, because "mulch is not used to keep the plants warm but to keep them frozen in the spring sunshine, by minimizing the damage caused by heaving of the soil."

What Ontario's own planting dates say, and what they don't

The province does publish hard planting dates, but for a purpose that is not yours. Appendix 2 of Ontario's AgriInsurance operational document sets the earliest and latest planting dates that commercial production insurance will cover, not what a gardener should do. Read with that condition attached, it is still the only official Ontario date list of its kind.

Two conditions govern every figure in it. First, the tables are split by insurance basis: crops "insured based on yield" carry a single planting deadline, while crops "insured on an acreage loss basis" carry earliest and latest dates in three crop-heat-unit classes: 2500-2700, 2700-3100 and over 3100. Sweet corn appears in both, at 5 July under the yield-based table and 25 June in the 2700-3100 acreage-loss class. That is not a contradiction; they are two different products, and quoting either as "Ontario's sweet corn deadline" would be wrong.

Second, and this is the limit that matters most here: the appendix never says which heat-unit class any Ontario municipality falls in. Searching the whole document for York, Peel, Durham, Halton, Toronto, Simcoe, "county", "region" and "municipal" returns four matches, all read: footnotes covering green peas and beets in Essex and Kent County, and garlic's coverage period. It does publish a "Map of planning deadline defined areas" and it does name regions in its footnotes, but those define Areas A to E for the grain, oilseed and fruit tables. It never assigns a crop-heat-unit class to any municipality, which is the assignment a GTA gardener would need. So compare the columns; do not assume which one covers you.

With those conditions carried, the useful figure is garlic's. Its latest insurable planting date runs 1 November in the 2500-2700 class, 5 November in 2700-3100 and 10 November above 3100: a spread of nine days across the entire province, which is why "plant garlic in the first week of November" survives as ordinary advice.