# Dubai residential sales: the DLD September filter capture, 30 September 2026

> A reproducible DLD residential-sales capture with exact filters, returned rows, registration and property-type cuts, hashes and revision limits.

Source: https://withlida.com/insights/dubai-residential-sales-september-filter-capture-2026

---

:::intro
The Dubai Land Department's 1 to 30 September filter returned **10,770 residential Sales rows worth AED 24,056,650,692.92** at 04:12:02 GST on 30 September 2026. This is a dated query capture, not a closed monthly total: 229 returned rows carry an `INSTANCE_DATE` of 31 August, and the latest returned timestamp is 29 September at 17:58:20.
:::

:::summary
- The official DLD response returned 10,770 rows and repeated `TOTAL: 10770`, so the 15,000-row request captured the complete response population at that cut-off.
- The rows sum to AED 24,056,650,692.92; the calculated median row value is AED 1,300,000.00.
- DLD labelled 7,258 rows Off-Plan and 3,512 Ready. Those groups account for 67.39% and 32.61% of the returned count.
- DLD labelled 9,442 rows Unit, 743 Building and 585 Land. These are property-type fields, not inferred home types.
- The filter boundary and the returned date field do not align exactly: 229 rows carry 31 August dates, so the result is reported as a 30 September capture rather than a final September total.
:::

## The capture returned 10,770 residential Sales rows

The most useful comparison keeps every figure inside the same official response. DLD returned the registration labels Off-Plan and Ready; the publication calculated their shares, sums and medians from the saved rows.

| Registration cut | Returned rows | Count share | Returned value | Value share | Median row value |
|---|---:|---:|---:|---:|---:|
| All returned rows | 10,770 | 100.00% | AED 24,056,650,692.92 | 100.00% | AED 1,300,000.00 |
| Off-Plan | 7,258 | 67.39% | AED 13,232,431,657.84 | 55.01% | AED 1,230,840.00 |
| Ready | 3,512 | 32.61% | AED 10,824,219,035.08 | 44.99% | AED 1,600,000.00 |

Source: [Dubai Land Department Real Estate Data](https://dubailand.gov.ae/en/open-data/real-estate-data/), official response captured 30 September 2026 at 04:12:02 GST; sums, medians and shares calculated from the saved rows.

The neutral read: Off-Plan supplied more returned rows and a larger share of returned value. The median is the middle `TRANS_VALUE` in each group. It is not a like-for-like property price because the response mixes Unit, Building and Land records.

::figure{src="images/blog/dubai-residential-sales-september-filter-capture-2026/inline-1.png" alt="Off-Plan and Ready row counts and count shares in the DLD 30 September 2026 capture" caption="Registration labels come from DLD; row shares are calculated from the 10,770-row response."}

The earlier [week-ending 24 September transaction pulse](/insights/dubai-residential-transactions-week-ending-24-september-2026) used a shorter filter window. Its population should not be added to this one because those rows already sit inside the wider query.

## The input register fixes the population before the total

The saved request sets Sales and Residential, then leaves registration type, freehold status, area and property type open. It asks for 15,000 rows from offset 0, sorted by the returned transaction-date field.

| Input or response field | Exact saved value | Treatment in this report |
|---|---|---|
| From and To | `09/01/2026` to `09/30/2026` | DLD's entered date-filter range |
| Transaction Type | Sales, `P_GROUP_ID=1` | Every returned row carries `GROUP_EN: Sales` |
| Usage | Residential, `P_USAGE_ID=1` | Every returned row carries `USAGE_EN: Residential` |
| Registration Type | Blank | Retains Off-Plan and Ready rows |
| Freehold, Area, Property Type | Blank | Applies no additional request filter on those fields |
| Page request | `P_TAKE=15000`, `P_SKIP=0` | Returned 10,770 rows; each row repeated `TOTAL: 10770` |
| Sort | `INSTANCE_DATE_DESC` | Orders rows by the returned field; does not define a separate population |
| Response time | 30 Sep 2026, 04:12:02.8149961 GST | Official capture cut-off |

Source: the official [DLD transaction interface](https://dubailand.gov.ae/en/open-data/real-estate-data/), its [public transaction script](https://dubailand.gov.ae/scripts/publicData.js?v=10) and the saved response from the POST endpoint `https://gateway.dubailand.gov.ae/open-data/transactions`, all accessed 30 September 2026.

The returned `INSTANCE_DATE` range is 31 August 2026 at 15:16:32 to 29 September 2026 at 17:58:20. Exactly 229 rows carry a 31 August date even though the From input is 1 September. The report preserves those official response rows. It does not replace DLD's boundary with a publication-created calendar cut or infer the gateway rule behind the difference.

That boundary is why the title says **September filter capture**. It does not call AED 24,056,650,692.92 an immutable September total.

## Unit, Building and Land are separate returned property types

DLD's property-type field divides the response into 9,442 Unit rows, 743 Building rows and 585 Land rows. The publication does not translate Unit into apartment, Building into villa, or Land into a named project.

| DLD `PROP_TYPE_EN` | Returned rows | Count share | Returned value | Value share | Median row value |
|---|---:|---:|---:|---:|---:|
| Unit | 9,442 | 87.67% | AED 16,511,954,812.09 | 68.64% | AED 1,194,914.00 |
| Building | 743 | 6.90% | AED 2,895,546,819.52 | 12.04% | AED 3,300,000.00 |
| Land | 585 | 5.43% | AED 4,649,149,061.31 | 19.33% | AED 3,809,280.00 |

Source: saved DLD response captured 30 September 2026; labels returned in `PROP_TYPE_EN`; all numeric columns calculated from the returned rows.

The neutral read: Unit rows account for 87.67% of count and 68.64% of returned value. Land has 5.43% of rows and 19.33% of value. Different record types and property mixes make the medians descriptive of this response, not a citywide home-price index.

The five `AREA_EN` labels with the largest row counts are shown without editorial renaming or identifier mapping.

| Returned DLD area label | Rows | Count share | Returned value | Median row value |
|---|---:|---:|---:|---:|
| JUMEIRAH VILLAGE CIRCLE | 789 | 7.33% | AED 1,062,567,895.97 | AED 960,000.00 |
| Madinat Al Mataar | 747 | 6.94% | AED 1,214,841,882.28 | AED 761,000.00 |
| MAJAN | 538 | 5.00% | AED 504,781,223.50 | AED 675,000.00 |
| BUSINESS BAY | 411 | 3.82% | AED 1,383,449,024.97 | AED 2,000,000.00 |
| Al Hebiah Fifth | 373 | 3.46% | AED 671,949,769.50 | AED 1,578,900.00 |

Source: saved DLD response captured 30 September 2026; names copied from `AREA_EN`; ranking, shares, sums and medians calculated from all returned rows.

An area label is not a project, building or community marketing name. Business Bay carries the largest returned value within this five-label sample, while JUMEIRAH VILLAGE CIRCLE has the largest row count. Neither fact establishes a price ranking because each label contains a different mix of returned records.

## Returned fields and publication calculations stay separate

DLD supplied the row fields, response timestamp and repeated total. The publication calculated the aggregate numbers and kept the raw response unchanged.

| Item | Record level | Returned or calculated | Use in the report |
|---|---|---|---|
| `TRANSACTION_NUMBER` | Response row identifier | Returned by DLD | Distinct-value check only |
| `INSTANCE_DATE` | Response row date and time | Returned by DLD | Returned range and boundary audit |
| `IS_OFFPLAN_EN` | Registration type | Returned by DLD | Off-Plan and Ready grouping |
| `PROP_TYPE_EN` | Property type | Returned by DLD | Unit, Building and Land grouping |
| `AREA_EN` | Area label | Returned by DLD | Exact-label grouping, no name mapping |
| `TRANS_VALUE` | Response row amount | Returned by DLD | Input to sums, averages and medians |
| Count share and value share | Group aggregate | Calculated by the publication | Group value divided by the full response total |
| SHA-256 | File or normalized row set | Calculated by the publication | Byte and population integrity checks |

The response has 10,770 rows but 10,767 distinct `TRANSACTION_NUMBER` values. The article therefore reports rows, not 10,770 unique legal transactions. The public response does not establish why an identifier repeats, so no cause is assigned.

The calculated average row value is AED 2,233,672.30, using AED 24,056,650,692.92 divided by 10,770. The average and median describe returned row amounts. They are not valuation opinions and do not merge this transaction table with DLD's separate rent, project, valuation, land, unit, building, broker or developer tables.

## The evidence bundle makes the capture repeatable

Reproducing this result means saving the input bytes, output bytes and calculation together. A later re-run answers a new as-of question because the registry population can change.

::figure{src="images/blog/dubai-residential-sales-september-filter-capture-2026/inline-2.png" alt="Three-stage route from the saved DLD query to raw JSON and cross-check" caption="The request, raw response and independent checks belong to one dated evidence bundle."}

| Saved artefact | Audit result | What it establishes |
|---|---|---|
| Exact query JSON | SHA-256 `f81475cc2a4e...` | The filter and page request used for this capture |
| Raw DLD response | 13,744,082 bytes; SHA-256 `c6df14cb7992...` | The exact official response used for every figure |
| Normalized returned rows | SHA-256 `c764eef53724...` | A content fingerprint excluding only page row number and repeated total |
| Independent calculation | All registration, property-type and freehold counts and values reconcile | Every displayed subtotal returns to 10,770 rows and AED 24,056,650,692.92 |

:::aside{title="Method and revision boundary"}
The calculation parses every `TRANS_VALUE` as decimal currency, then recomputes row counts, sums, averages, medians, shares and group checks from the saved response. The official DLD page publishes no formal revision calendar or finalisation boundary, accessed 30 September 2026. This article is therefore a **30 September 2026 capture**. The [26 September query-revision test](/insights/dubai-transaction-query-revisions-26-september-2026) shows why a later run should be labelled as a later snapshot rather than silently substituted.
:::

The broader [DLD real-estate data source map](/insights/dubai-dld-real-estate-data-sources-2026) keeps transaction, rent, project, valuation, land, unit, building, broker and developer routes separate. This capture uses only the transaction response.

## FAQ, checked 30 September 2026

::::faq
:::faq-item{q="How can I check Dubai real estate transactions?"}
Use the official [DLD Real Estate Data](https://dubailand.gov.ae/en/open-data/real-estate-data/) service. Record the From and To dates, transaction type, usage and every open filter, then save the response timestamp with the result. The page was accessed 30 September 2026.
:::

:::faq-item{q="Is AED 24,056,650,692.92 the final September 2026 residential-sales total?"}
No. It is the sum of the 10,770 rows returned at 04:12:02 GST on 30 September by the exact filter shown above. The official page publishes no formal revision boundary, and the returned date range includes 229 rows dated 31 August.
:::

:::faq-item{q="Why did a 1 to 30 September DLD filter return rows dated 31 August?"}
The official response returned 229 such rows. This report preserves that boundary and does not infer a timezone, registration or gateway rule that the public response does not state.
:::

:::faq-item{q="Are 10,770 rows the same as 10,770 unique transactions?"}
No. The capture contains 10,770 rows and 10,767 distinct returned `TRANSACTION_NUMBER` values. The report uses row language because the public response does not establish why some identifiers repeat.
:::

:::faq-item{q="What does Off-Plan mean in this table?"}
It is DLD's returned `IS_OFFPLAN_EN` registration-type label. It is not treated here as a separate statement about construction progress, project completion or unit availability.
:::
::::

:::callout{variant="info" title="Subscribe for the next quarterly report"}
Subscribe to the withlida market-data briefing for the next quarterly report and its dated source register.
:::