How late is late? What MahaRERA's possession delay data shows

Of 56,061 published MahaRERA registrations that carry both dates, 12,689 now promise possession later than they first did. That is 22.6 percent of the register, about one project in five. Among them, the typical revision is 21 months, and one in four has moved its date by more than three years. MahaRERA's possession delay data is unusually honest for one reason: the regulator keeps the first promised completion date beside the current one, so a delay can be measured without anyone having watched it happen.

The question this data answers#

A buyer looking at an under-construction flat is buying a date as much as a flat. The question is how far that date tends to move. National studies have measured the backlog: ANAROCK's 2021 analysis counted about 6.29 lakh homes, worth about Rs 5 lakh crore, delayed or stalled across India's seven largest city regions. The MahaRERA register lets the question be asked project by project in Maharashtra, because every registration files an original proposed completion date and a current one. Where the current date is later, the project has revised; the gap is the revision.

Two cautions shape every chart below. First, a project cannot revise a date it has not yet approached, so young registrations look punctual simply because they are young. Second, these are the dates as filed. MahaRERA's four month extension of 7 August 2026 applies automatically to eligible projects, but as updated on 3 October 2026, 17,227 of the 18,524 projects whose first date falls on or after 28 February 2026 still show that first date in their filing, so the figures here largely leave it out.

How often dates move, by the year of registration#

The 2017 cohort stands apart because it includes the projects already under way when RERA began, which registered all at once with dates set under the old regime; 5,678 of its 14,736 projects have revised. For registrations from 2018 to 2023, whose dates have mostly arrived, the share settles at about one in five. The older cohorts also lived through the pandemic, when MahaRERA extended completion dates for qualifying projects by orders of 2 April 2020, 18 May 2020 and 6 August 2021; where a promoter's filing carries those extensions, they count as revisions here. The falling share from 2024 onwards is age, not improvement: most of those dates are still ahead.

How far they move#

The single commonest revision is exactly 12 months (2,893 projects), followed by 24 months (1,163) and 18 months (895). Round numbers are a clue to how revisions happen: a promoter applies for an extension of a year, or of two, rather than for the time the building actually needs. Section 6 of the RERA Act allows the Authority to extend a registration for force majeure by up to one year in aggregate, which is part of why a year is the unit of choice. Note the tail: a quarter of revised projects have moved by more than three years, and those are the ones a buyer most needs to see before booking.

Where dates move most#

Mumbai's two districts and Palghar revise most often, and Palghar's revisions run longest. Read the comparison with care: districts differ in how old their projects are and how large, and both affect how often a date moves. The chart says where revisions are common, not which builders or buildings are slow.

What happened when the first date came#

About half of the projects whose first date has passed read Completed, and only 582 Active projects are past even their current date. The large Lapsed share needs the most care. A registration lapses when its date passes without a completion filing or an extension, and 9,653 of the 13,932 Lapsed projects never filed a later date at all. Lapsed is not the same as abandoned. When MahaRERA served notices on about 10,773 lapsed projects in December 2024, developers of 5,324 had responded by early 2025; 3,517 of them submitted occupancy certificates, 524 asked for extensions, and the registrations of 1,950 projects were suspended with their bank accounts frozen. Many lapsed projects were built and simply never closed on the register; others had stopped. The status alone cannot tell you which, as lapsed, revoked, deregistered explains.

An analogy: the arrival board#

A railway arrival board shows two times for a late train: the time printed in the timetable and the expected time now. The gap tells you how late the train is running today. It does not tell you when it will arrive, and a train already an hour late often loses more time before the platform. A project's two dates work the same way. The revision is a measure of the delay so far, and the construction progress in the quarterly filings is the only evidence of whether the new date will hold.

Priya and Arjun compare two dates#

Priya and Arjun (illustrative, as our stories always are) had two Pune projects left on their list. The first had promised December 2025 and now promised December 2027, a 24 month revision with an extension certificate on file and a stated reason. The second, registered in 2025, had never revised and was not due until 2028. Their spreadsheet treated neither date as a fact. For the first, they read the quarterly progress filings to see whether construction had actually moved since the extension; for the second, they read the promoter's earlier registrations, because a young project's clean record says nothing yet. Names and numbers in this story are illustrative.

What this means for you#

• Read both dates, not one. The registration shows the first promised completion date and the current one; a gap between them is the project's own record of delay.
• Ask why it moved. A revision backed by an extension carries a stated reason on the filing. Extensions under MahaRERA and force majeure in RERA explain which reasons hold up, and what RERA extension data says counts the reasons promoters actually give.
• Do not trust youth. A new registration has not had the chance to slip. The promoter's record across its earlier projects, read through checking a builder's track record, is the better forecast. Comparing a young project fairly against older ones means dividing by its months on the register, which comparing flats before buying does for a shortlist.
• Know which date is yours. Delay rights run from the date in your registered agreement, as RERA rules for delayed possession and computing delay interest explain.

What the report does with the dates#

The Rs 499 ReraGenie buyer report tells the possession story of one project: the first promised date against the current one and the gap between them, how the promoter's other registrations moved their dates and how each ended, and the project's construction pace measured against other recent projects in the same district. It closes with your rights under RERA, including a worked delay-interest example. If you have already booked, the project watch that comes free with the report emails you when the date or the filing changes. Every project's free page on ReraGenie draws both dates on a possession timeline, so a shortlist can be screened before anything is paid.

Methodology and sources#

• Date figures: ReraGenie analysis of 56,068 published MahaRERA registrations, as updated on 3 October 2026; 56,061 carry both an original and a current proposed completion date. A revision is a current date later than the original, measured in whole months; 31 projects whose current date is earlier are excluded from the revision counts.
• Status figures use the registration status on the same date. The four month extension of 7 August 2026 is reflected in the filed dates of few eligible projects, so the counts largely exclude it.
• MahaRERA notices on lapsed projects: figures released by MahaRERA in early 2025, as reported at the time.
• National backlog: ANAROCK Research, 2021.
• Pandemic extensions: MahaRERA orders of 2 April 2020, 18 May 2020 and 6 August 2021. Law: Real Estate (Regulation and Development) Act, 2016, sections 6 and 18.