The Best Trick-or-Treating in Vancouver, Ranked: 22 Neighbourhoods by Doors, Lights and Sidewalks

Kensington-Cedar Cottage packs 86 residential lots onto every kilometre of street. Downtown has 0.2. If you're filling a pillowcase on October 31, that difference is the whole game.

I wanted a straight answer to a question every Vancouver parent argues about: where's the best trick-or-treating in the city? So I scored all 22 of the City of Vancouver's local areas on three things the City publishes as open data. Doors: how many residential lots line each kilometre of street, because more front doors per block means more candy per step. Light: street light poles per kilometre. Sidewalks: how much of the sidewalk the City rates in good or very good condition, because nobody wants a Spider-Man face-first on a heaved slab.

Then I pulled every October 31 on record at Vancouver International Airport, back to 1937, to answer the other question: will it rain? (Spoiler: bring the umbrella.)

The map: where Vancouver trick-or-treats best

How the score works: I ranked the 22 areas on each measure and turned the rank into a 0 to 100 scale (best = 100, worst = 0), so one extreme value can't swamp the rest. Then I blended them: doors count for 60%, light and sidewalks 20% each, because no amount of street lighting helps if there's nobody to knock on. That's my weighting, not a law of nature, so the sliders let you set your own. A parent of a three-year-old might care most about sidewalks. A twelve-year-old with a plan cares about doors and nothing else.

Source: City of Vancouver Open Data (street lighting poles, public streets, sidewalk condition rating, property parcels, zoning districts), pulled Oct 13, 2026. Contains information licensed under the Open Government Licence – Vancouver. Analysis by EstateBlock. Nicknames are ours and meant as a joke. Neighbourhood links go to the closest EstateBlock neighbourhood page; the boundaries don't match the City's local areas exactly.

The east and south side sweep the top of the list: Kensington-Cedar Cottage, Hastings-Sunrise, Renfrew-Collingwood, Victoria-Fraserview and Sunset, all scoring between 69 and 73. My read: these are blocks of houses packed close together, with short walks from one porch to the next. Kensington-Cedar Cottage wins on doors (86.2 residential lots per km, the most in the city), with middle-of-the-pack lighting (39.4 poles per km) and 70% good sidewalk. It's close at the top, though: move the sliders a couple of notches and Hastings-Sunrise or Renfrew-Collingwood takes over.

The bottom of the list is the West Side. Kerrisdale comes last (26) and Shaughnessy second-last (31). They have the dimmest streets in the city (25.6 and 26.1 poles per km) and fewer doors than average: Shaughnessy has 39.4 residential lots per km and Kerrisdale 52.5, against 61.1 city-wide. In Shaughnessy you will walk a long way between doors, in the dark, and some of those doors are at the end of a very long driveway. Shaughnessy's sidewalks are lovely (85% good), which helps less than you'd think.

And then there's Downtown, #18 of 22. It has by far the most light (90.5 poles per km, 1.7 times the next-brightest area) and the best sidewalks (88% good), but only 11 residential-zoned lots across 57 km of street. Most of Downtown is zoned for towers, mixed use and the historic districts, and a tower is one lobby with a concierge who is not handing out Coffee Crisp. The West End has a milder version of the same problem: 32.7 lots per km, most of them apartment buildings.

A caution on the door count: a lot is a lot, whatever sits on it. A house with a suite and a laneway home is one lot with up to three doors. A four-storey rental building is one lot with one buzzer. So this measure rewards exactly what trick-or-treaters want: lots of separate front doors at street level. If you sort by doors alone, the order goes Kensington-Cedar Cottage, Sunset, Victoria-Fraserview. If you ignore doors and only score light and sidewalks, Downtown wins, followed by Fairview and West End. Which is a nice reminder that a well-lit sidewalk with nobody handing out candy is just a sidewalk.

The City has 58,008 street lighting poles in its inventory, about 40 for every kilometre of street. Kerrisdale (25.6), Shaughnessy, West Point Grey and Dunbar-Southlands sit around 26 per km, roughly two-thirds of the city average. This is a pole count, not a brightness reading: it doesn't know about porch lights, jack-o'-lanterns or the neighbour who goes full haunted house. But on a dark, wet October night it's a fair proxy for how easy it is to see a small ghost crossing the road.

The City rates the condition of 2,055 km of sidewalk. 75% of it is good or very good and 12% is poor or very poor. The best concrete is in Downtown, Fairview and West End. Sunset has the most to complain about, with 62% good and 23% poor or very poor. Not a reason to skip it (it's #2 on doors alone), but maybe not the year for the floor-length witch costume.

Will it rain on Halloween in Vancouver?

Probably. In the last 30 years (1996 to 2025), Vancouver International Airport recorded measurable rain on 20 of 30 Halloweens. 13 of them got 5 mm or more, which is the "costume under a raincoat" level. 10 were completely dry. Over the full record back to 1937 it's 56 wet Halloweens out of 89, so this isn't a new thing.

The wettest Halloween on record was 1981, with 57.2 mm. The warmest afternoon was 1949 at 16.1°C, and the coldest was 1984, when the high only reached 1.2°C and it dropped to -5.9°C. No snow has been recorded at the airport on any October 31 in the 89-year record. The typical Halloween of the last 30 years topped out around 11°C, and 6 of them dipped below zero overnight.

One caveat: these are daily totals, so a morning shower that cleared by 5 pm counts as a wet Halloween. The airport is also on Sea Island in Richmond, not in your neighbourhood. But rain in Vancouver in late October is rarely a one-neighbourhood event.

Source: Environment and Climate Change Canada, daily climate data for Vancouver International Airport (climate IDs 1108447 for 1937 to 2012 and 1108395 for 2013 to 2025). Analysis by EstateBlock. "Wet" = 0.2 mm or more of precipitation on the climate day.

What this has to do with buying a home

For the numbers that actually move prices, see my Greater Vancouver market update or my take on whether it's a good time to buy a house in Vancouver. If you want a neighbourhood's personality instead, my 3-1-1 complaints map shows what each one reports to the City, and when it snows in Vancouver covers the next holiday's weather. To look at houses on these streets, there are 1,330 Vancouver houses for sale on EstateBlock right now (as of 2026-10-13), or browse Knight, Main and Collingwood.

How I did this

Everything comes from the City of Vancouver's open data portal, pulled on October 13, 2026, aggregated to the City's 22 local areas:

• Doors: the City's property parcel polygons (99,698 parcels), each placed in a local area by its centre point and in a zoning district from the City's zoning map. I counted parcels in districts the City classifies as Residential, Residential Inclusive or Residential Rental, plus the First Shaughnessy District, which the City files under Comprehensive Development but is a house zone (88,047 in total), and divided by the kilometres of City street in each area from the public streets dataset (1,441 km). Other Comprehensive Development districts (CD-1 sites, the Downtown District and others), Historical Area, commercial and industrial lots are left out, which is why Downtown scores so low on doors.
• Light: every pole in the street lighting poles dataset, using the City's own local area field (303 poles have no area and are left out), divided by street km.
• Sidewalks: the sidewalk condition rating dataset, measured by length. Each block face is placed in a local area by its centre point.
• Weather: Environment and Climate Change Canada daily climate records for October 31 at Vancouver International Airport, 1937 to 2025 (two station IDs for the same airport: 1108447 until 2012, 1108395 since 2013).

Things I left out on purpose: crime and collision data. Halloween is supposed to be fun, and a neighbourhood's police calls say more about where people and businesses are than about whether a kid gets a good haul. Nothing here measures who actually hands out candy, or how generous they are. That's still a field-research problem.