Methodology

Every fact here traces to a public source. Where we cannot source something, the page says so rather than guessing.

What is covered

Sources

Program details — name, address, phone, coordinates, capacity, age groups, licence status and $15/day participation — come from the Government of Alberta's childcare search.

Inspection history comes from the Child Care Information dataset, published quarterly under the Open Government Licence – Alberta. It is published with its own usage considerations, which we have read and which shape several of the statements on this site. These two provincial sources use different program identifiers with no overlap, so records are joined on contact details instead — see reading inspection findings for the rule and what we do when it cannot be met.

Schools served comes from two sources that say different things, and every link on a daycare's page cites which one it came from.

Most come from each Edmonton Public school's own childcare page, which exists in no government dataset. Pages are prose, so daycares are extracted and matched back to licensed programs by phone, name and address; only high and medium confidence matches are published. A claim that contradicts the measured distance — a daycare described as on-site two kilometres away — is withheld for review.

The other 55 come from the province writing the school into the program's own licensed address: “9735 80 AVENUE MILL CREEK SCHOOL”. That is the licensing record placing the program in the building, which is a narrower claim than a school naming a program as serving its pupils, so we publish it as a location and nothing more. We accept it only when every distinctive word of exactly one school's name appears in the address — an ambiguous name is refused rather than guessed — and when the two sets of coordinates agree to within 300 metres, which is an independent check: the geometry played no part in making the match. One program was refused for ambiguity and one for sitting six kilometres from the school its address named.

Being in the building is not the same as serving the school, and the licence type does not close the gap: an out-of-school-care licence says a program may run out-of-school care, not that it takes children from the school it rents a room in, still less that they walk there. 11 of these programs hold no out-of-school-care licence at all — a preschool renting a room, or a daycare for the parents who attend the school. So an address never carries a claim about how children travel, and both the daycare and school pages show these under their own heading, as tenancy, apart from the programs a school has named itself.

Neighbourhoods are assigned by point-in-polygon against the City of Edmonton's 407 neighbourhood boundaries, so they reflect where a daycare actually sits rather than how its address is written. One program near the southern city edge falls outside every published boundary and shows none.

Opening hours come from the operator's own website and appear in no government dataset. We keep a claim only where one sentence carries both the days and both times, and we publish that sentence beside the hours with a link to the page it came from. An operator running several locations often publishes one set of hours for all of them, so the hours are what their site says rather than a promise about this address.

Websites appear in no government dataset, so each one is found by us and then has to earn its place. A site is published only when the page carries something of the program's own — its phone number, its postal code, or its street address. A domain that merely resembles the name is not evidence and is not published; see how we keep one daycare's website off another's page.

Maps are rendered live by Google's Maps Embed API. We do not download, cache or re-host Street View or satellite imagery.

What this cannot tell you

• Out-of-school care and preschool fees. Alberta's $15/day grant covers daycare only. Those fees are set by each operator and are not published anywhere.
• Optional fees — meals, field trips, transport, supplies — vary by operator and are not in any dataset.
• Whether a family qualifies for the income-tested subsidy. Alberta runs a separate subsidy alongside the $15/day grant, aimed particularly at kindergarten and school-age children in out-of-school care. Eligibility depends on household circumstances we neither hold nor want, so we point to it rather than estimate it.
• Availability. Nothing here tells you whether a program has room. The province's data carries a field called Enrollment, and we used to publish it and subtract it from capacity — until a reader asked why one program showed 69 children enrolled against 56 licensed spaces. It is not a count of children: 93% of Edmonton programs report it exactly equal to their licensed capacity, and only 3% below. Where it differs, it usually matches the number the province's other file calls capacity — of the 32 programs that appeared over-enrolled, 30 had an inspection-file capacity equal to this figure. So it measured no one, and we no longer publish it or anything derived from it. Call the operator to ask about openings.
• Most school links. 173 of 554 out-of-school-care programs have a confirmed school, and only 123 of the 212 links state whether children walk or are driven. Catholic and independent schools are not yet covered.
• Websites and photos. The provincial data carries neither. Street View shows the building; operator photos are added only with permission.
• Quality. We publish no ratings, scores or rankings. Inspection findings are shown as the province records them, with their remedy dates.

Reading inspection findings

Findings are routine in licensed childcare and most are remedied quickly. Two states that look similar are not the same, and we keep them apart: 92 programs have no inspection in the published window — often because they are newly licensed — while 385 have been inspected with no findings. An absence of records is not evidence of a clean record.

That window matters, and it is narrower than it looks. The province's own usage considerations say this file “comprises a point-in-time data pull” holding “18 months of data on open child care programs that have an active license during the last month of the quarter”. The file we publish from runs July 2024 to December 2025. So every count on this site is a count within those months, and an empty history means no inspection in that window — never that a program has never been inspected. Programs that closed mid-quarter are not in the file at all.

Two more things the province says, which we pass on rather than paraphrase away. A program has 30 days to appeal a non-compliance it considers unwarranted, so a finding shown here may since have been overturned. And the file is revisable: rows for the same period “may not necessarily be identical from one post to the next due to data entry modifications and/or program activity”. Neither is a reason to hide findings; both are reasons to read one as a dated record rather than a verdict.

Every finding's date links to the Government of Alberta dataset it was read from, so you can check the row yourself rather than taking this site's word for it.

The province publishes inspections in one file and program details in another, and the two share no common identifier. We join them only when two independent identifiers agree — phone with name-and-postal, or name-and-postal on its own — and we refuse the join outright whenever they point at different programs. There is no score or threshold that lets a near-match through. A phone number on its own is not enough, because an operator running several centres often publishes one head-office number for all of them. Findings are attributed to the single licensed program the province issued them against, and are never rolled up to an operator or a chain. Where a row cannot be resolved to exactly one program, it is left out entirely.

That rule is then checked with things it never used. The inspection file carries its own address, its own capacity figure, and its own daycare / out-of-school-care / preschool flags, none of which take any part in the matching — so they can confirm a result rather than restate it. On the current snapshot the street number agrees on every single one of the 4,148 joined rows where both files give one, and the service flags agree on 96%. Capacity agrees on 89%, and the disagreements are concentrated in 79 programs whose address and phone match exactly — day home agencies for the most part, whose roster of homes grows and shrinks between the two files' publication dates.

One failure could survive all of that. The inspection file has an identifier of its own — useless for joining, since it shares nothing with the province's program numbers, but stable enough to say which rows belong to one licence. If two licences were ever to land on the same program, every row would still look correctly matched on its own while two different daycares' findings sat on one page. So that is checked too, across the whole file rather than row by row, and a program drawing findings from more than one licence keeps none of them. It refuses nothing today; it is there so that it cannot start happening quietly.

Keeping one daycare's website off another's page

Publishing the wrong operator's website would be worse than publishing none, so the rule is a positive one: a site is shown only where we can point at something on the page that belongs to this program. Phone number, postal code or street address — any one will do, all three are better, and a name that merely looks like a domain does not count.

That bar was not always enforced. 77 sites had been accepted on a name resemblance alone, so every published site was re-read from scratch and tested against a corpus it had never been matched on: all 3,039 licensed childcare programs in Alberta. The question asked of each page was not “does this look right” but “whose contact details are actually on it”.

The sites that had been corroborated when they were found held up: re-read, 87% still carried this program's own phone or postal code, and 7 of 480 carried another program's and none of their own — each of those an operator's own site naming a sister location, or an office shared with a second agency, and each checked by hand and recorded with its reason. The sites accepted on a name did not hold up. Only 16 carried anything of the program's; 13 carried some other licensed program's instead. Two were plainly a different business: one domain belonged to a dayhome agency operating under another name entirely, and another daycare's page had been given the website of a company it has no connection to.

So the uncorroborated ones are no longer published. That cost 77 links, and some of them were surely right — a chain's own website, dropped because we could not tie it to the particular location whose page it would sit on. It is the right way round. A missing link is a gap a parent can work around; a wrong one is a false statement about somebody's business, printed beside their inspection record.

Family day home agencies

34 of the licences listed here belong to family day home agencies rather than childcare centres. An agency places children with contracted educators who care for them in their own homes, so its listed address is an office, the province records its licensed capacity as zero, and the home a child would actually attend does not appear in the provincial data at all. Where such an agency has a non-compliance finding, that finding is recorded against the agency's licence. The province is explicit about what that covers: non-compliances issued to a family day home agency “reflect Agency non-compliances as well as family day home provider” ones. So such a finding may be against the agency's own practice or against one of the homes it approves, and the file does not say which, or how many homes are involved. Those pages say so, and we do not repeat the zeros as though they described space.

Corrections

If you dispute a finding shown against your program, we take it down first and check it afterwards. Tell us which record and we withhold it from the page on the next deploy — usually the same day, and within three business days at the outside — while we compare it against the provincial source. If the record holds, it goes back up and we tell you why. If it does not, it stays down. Every withheld record is counted on the affected page, so nothing disappears silently. No records are withheld today.

If you operate a program listed here and something is wrong, missing or out of date — fees, schools served, your website, photos you are happy for us to host — please get in touch and we will correct it and note the date. Provincial licensing data itself has to be corrected with the Government of Alberta; we re-import it on a schedule and your changes will flow through.

What we have corrected

Nothing has been corrected yet — no one has told us we were wrong. When someone does, it gets logged here, including the times we checked and the record held.

Data rebuilt 2026-08-30. This site is not affiliated with the Government of Alberta, Edmonton Public Schools, or any operator.