Central Allocation choice form: Part A and Part B strategy (SSPA)

What this page answers

This page covers only the 2025/2027 Central Allocation choice form: how to fill up to 3 Part A choices and up to 30 Part B choices, the order the computer uses, and what to check on the day. Students now in Primary 6 (September 2026) enter Secondary 1 in September 2027 and are still scored 999632. Discretionary Places windows, successful-applicant notices and matching rules are on theSSPA processpage. This page assumes the family has reached the April Secondary One Choice of Schools form.

How the form is structured

After deducting repeaters’ places and Discretionary Places, the remaining Secondary 1 places are allocated by computer: 10% in Part A (unrestricted by school net), up to 3 choices in any net including the student’s own; 90% in Part B (by school net), up to 30 choices in the student’s own net. There are 18 secondary school nets, matching the 18 administrative districts. In principle a student’s net is the district where their primary school stands. Each net includes that district’s participating schools plus some other-district schools that offer places into the net; the number varies by year.

Own-net schools may be repeated in both parts; the same school must not be repeated within one part. The Education Bureau warns that listing only a few favourite schools in Part B does not raise the chance of getting them. The e-platform prompts when Part B has fewer than 20 choices. Fill as many of the 30 own-net choices as possible. The Notes for Parents on Central Allocation are uploaded on 8 April 2027; the Secondary One Choice of Schools form and the secondary-school list (《中學一覽表》) are issued in early April 2027. The e-platform window is 8 April to 5 May 2027; paper forms go back to the primary school in early May. Whether or not Discretionary Places was applied for, this form must be returned.

How the computer processes choices

Allocation uses three factors: band, the parents’ choice order, and a random number. The computer first standardises each student’s three internal-assessment terms, then adjusts them by the average of the primary school’s two most recent Pre-S1 HKAT samples. For the 2025/2027 cycle those samples are 2024 and 2026. Adjusted scores are cut two ways: a territory band that splits all students into three equal thirds, used for Part A; and a net band, cut inside each of the 18 nets, used for Part B. Bands are computed only when the allocation programme runs and are never released to parents.

The order of processing is fixed. The computer handles Part A first: every territory Band 1 student’s up to 3 choices from the top, then territory Band 2, then territory Band 3. Anyone placed in Part A is not processed in Part B. Then Part B: net Band 1 in choice order. A Band 1 student whose 30 Part B choices are all full is assigned a place from the net’s remaining places; then net Band 2, then Band 3. The random number is generated before the run. It only orders students in the same band when a school is over-subscribed (smaller number first). It is unrelated to the student number or personal data.

Part A strategy

This section is general guidance, not an Education Bureau rule. Part A is only 10% of the remaining places and is processed by territory band: every territory Band 1 Part A choice is considered before Band 2 or Band 3, and the field is every applicant, not just the own net. That makes Part A the only box in which a school outside the student’s net can be written. Using all three rows to repeat “safe” own-net schools that will also appear in Part B gives up the unrestricted window; those own-net schools can be listed again in Part B. A common parent habit is to keep Part A for schools the family genuinely wants that sit outside the net, or for a long-shot that is still worth one row. That explains the mechanism. It does not promise a place.

Part B strategy

This section is general guidance, not an Education Bureau rule. Part B is 90% of the remaining places, own-net only, and runs by net band from high to low, then by the parents’ order. Inside the same band the computer walks choice 1, then 2, then 3. A school written lower is not “easier to get”. Order should match genuine preference: first the schools the family most wants, then those it can accept, then fallback schools. Some families split the 30 rows into favourite, solid and safety blocks. That is a parent habit, not a Bureau format. Blanks do not raise the chance of a favourite. The Bureau asks families to fill as many of the 30 as possible; the e-platform prompts below 20 rows.

If the primary has a feeder or nominated secondary, reserved places have extra conditions. After repeaters and Discretionary Places, a feeder secondary reserves 85% of the remainder for its feeder primary; a nominated secondary reserves 25%. Eligible students are those in net Band 1 or 2 who put that school as their Part B first choice. Listing it as Part B choice 2 or later does not use the reservation. Cross-net approval forfeits reserved-place eligibility. Unfilled reserved places go into open allocation.

A worked example

A fictional student, Lok (小樂), shows how one 30-row list is processed. School names are omitted and no probabilities are stated. She is in Primary 6 in 2026/27. Her net is the administrative district of her primary. The family used this site’sbanding estimatorand obtained a net-Band-2 estimate. The actual band is computed only when the programme runs and is never released; the estimate may differ from the result.

In Part A she lists 3 schools in other nets. Her 30 Part B rows follow genuine preference: favourites first, then acceptable schools, then fallbacks. The computer runs Part A first. If Lok’s territory band is 2 or 3, all territory Band 1 Part A choices are processed before hers. When her turn comes, the computer checks her Part A list from choice 1. If one of those schools still has a place, she is allocated it and the whole Part B list is unused. If all 3 are full, she enters Part B. Part B first clears every net Band 1 student. Only then, in net Band 2, does it walk Lok’s Part B choices from 1, then 2, then 3. When the same school is over-subscribed inside the same band, the smaller random number is considered first. If she has a feeder school but did not put it as Part B first choice, that school’s reservation does not apply to her.

Checklist for the day

The Secondary One Choice of Schools form and the secondary-school list are issued by the primary in early April 2027; the Notes for Parents are uploaded on 8 April. Before filling, check each school code and address on the list, and confirm the school belongs in the right part: Part A may cross nets; Part B must be own-net. Do not repeat the same school inside one part. An own-net school that should appear in both parts must be written once in each.

On the eSSPA e-platform the window is 8 April to 5 May 2027. After signing with iAM Smart+, complete “confirm and submit” (確認及提交); otherwise the form is not filed. Paper forms go back to the primary in early May. Families notified as Discretionary Places 正取 on 31 March only tick the box, strike out the choices section and sign. Filling choices anyway does not change the outcome: a student allocated a Discretionary Places offer receives no Central Allocation place.

Common misunderstandings

“Listing fewer Part B schools makes a favourite more likely.” It does not. The Bureau warns that this does not raise the odds; the e-platform prompts below 20 Part B choices.

“Part A is a free extra chance and does not affect Part B.” It is not. Part A is 10% of the remainder and is processed before Part B by territory band. A Part A allocation stops Part B. It is not a second insurance policy on top of Part B.

“Schools can see which choice number I gave them.” At Discretionary Places, parents keep the fourth copy that shows the order; schools do not see the ranking. At Central Allocation the computer uses the order to allocate inside a band. Individual schools do not mark the form as an interview score.

“A 31 March 正取 notice means we need not return the Central Allocation form.” The form must still be returned. Successful families only tick the box, strike out the choices and sign. Results are announced with Central Allocation on 6 July 2027.

“Leaving a few Part B rows blank does no harm.” The Bureau asks families to fill as many of the 30 own-net choices as possible. A Band 1 student whose 30 choices are all full is assigned a remaining place in the net. Blanks do not raise the chance of a favourite.

Next steps

For how Discretionary Places joins Central Allocation, see theSSPA process. For an estimate of banding (not an actual band), see thebanding estimator. For every date in this cycle, see theSSPA timeline.

FAQ