Skip to content
Utah Community Learning

Frost dates are a suggestion, not a promise

About 20 minutes

Frost dates are a suggestion, not a promise

Every seed packet and every extension office chart will give you a last frost date for the valley, usually somewhere around the first two weeks of May, and a first frost date in fall, usually late September or early October. Write those down. Then don't trust them completely. That's really the whole lesson, but let me explain why, because it matters for everything else we do this module.

A frost date is an average. It's built from years of weather data smashed together into one number, which means half the years on record froze after that date and half didn't. Utah County sits at the base of the mountains at around 4,600 feet, and cold air rolls down off the canyons at night and pools in the low spots. Your yard might be three degrees colder than the airport where they measure the official date. Or three degrees warmer, if you're up on a slope where the cold air drains away. Nobody's chart knows your yard. You have to.

Here's what actually happens if you plant to the calendar instead of the weather: you get a warm stretch in late April, everyone's outside, the garden centers are slammed, and you decide you're clear. I did exactly that my first year. Planted tomatoes the second week of May because it felt like summer had arrived. Third week of May, hard frost, killed every plant I had. I replanted, and I taped a frost calendar inside my shed door that week. It's still there. I don't fully trust a warm April afternoon anymore, and honestly I don't think you should either.

What to actually do

Keep a real frost log for your own yard. Not the county average, yours. Note the date of first and last frost each year, right on a calendar or a notebook by the back door. After three or four years you'll have a better number than any chart, because it's specific to your dirt, your slope, your trees.

Watch the actual forecast, not the season. Once you're within a couple weeks of your average frost date, check nightly lows every evening. A clear, calm night is the dangerous one — clouds hold heat in, clear skies let it radiate straight out. If it's clear and the low's dipping toward freezing, that's your warning, whatever the calendar says.

Keep cover ready even after you think you're clear. Row cover, an old sheet, a bucket over a pepper plant. I keep a stack of row cover folded in the shed from April through early June, because the valley has burned me on a "surely we're past it" night more than once.

Don't rush the tender stuff. Tomatoes, peppers, squash, beans. They don't just dislike frost, they sulk in cold soil even without one. Planting them two weeks early doesn't gain you two weeks, it just stresses the plant and you end up even.

Do push the hardy stuff. Lettuce, spinach, carrots, peas can handle a light frost, some of them handle it well. This is where you actually gain real weeks, and it's a big part of why we're spending this whole module on season extension instead of just planting later.

This connects to an opinion I hold pretty firmly: frost dates are a suggestion, not a promise. Plant to the weather in front of you, not the number on the packet.

Speaking of learning by getting it wrong

My daughters, Maggie and Mia, have their own spot in one of the raised beds. Mostly carrots, and one pumpkin plant that's been weirdly determined every year we've grown it. Half of what they plant doesn't come up, and I've made peace with that a long time ago. They're learning the same way I did, which is to say: plant it, watch what happens, get surprised, try again next year with a little more information than you had. Frost dates are the same deal, just with higher stakes than a carrot that never sprouted. You're going to lose something to a late frost at some point no matter how careful you are. That's not a failure of the system. That's just the valley.

Before next time: if you don't already have one, start a simple frost log this week — even just a sticky note on the fridge with this year's first warm stretch and whatever the overnight low actually does. You'll want that data once we get into cold frames.