Prolepsia forecastNumber
How many Starship launches will SpaceX make between 1 January and 31 December 2027?
As of 17 November 2026, Prolepsia’s median forecast is 14.1, with an 80% chance it falls between 6 and 30.
14.1median
How Prolepsia reasons
The report Prolepsia wrote with this forecast.
How forecasts are made · Track record
TL;DR
The forecast for January 1–December 31, 2027 (UTC) is a median (50th percentile) of 14.1 launches, a 10th percentile of 6 launches, and a 90th percentile of 30 launches. This anticipates substantial growth, but not the routine weekly operations SpaceX is targeting.
Key evidence
- The historical starting point is modest. Integrated launch counts were 2 in 2023, 4 in 2024, 5 in 2025, and 4 through November 17, 2026. The 2026 flights occurred on May 22, July 24, September 28 and November 13, with calendar-date gaps of 63, 66 and 46 days (launch records).
- Recovery progress is real, but incomplete. Flight 15 deployed 26 operational Starlink V3 satellites and recovered the upper stage after approximately 22.5 hours. The booster aborted its catch after an early boostback shutdown associated with low fuel mass and was destroyed offshore. This supports acceleration without establishing repeatable full-system reuse (SpaceX postflight report).
- Capacity is not throughput. Starbase's existing allowance is 25 annual launches; larger proposals remain under review. LC-39A's environmental envelope covers 44 annual launches, but environmental clearance is distinct from an operating license (Starbase FAA materials, LC-39A FAA materials). Neither figure is a worldwide launch forecast.
- Demand supports growth without guaranteeing a schedule. Satellite deployments already provide useful work. NASA's revised plans place Artemis III's Earth-orbit demonstration in 2027 and lunar landing objectives in 2028, so a complete lunar-refueling campaign cannot simply be assumed for the forecast year (NASA update).
Uncertainties
The decisive unknowns are qualified vehicle production, booster recovery reliability, actual refurbishment time and Florida's operational readiness. A new pad adds little worldwide output if it draws from the same limited supply of flight-ready vehicles.
A failed flight still counts, but its investigation or repairs can prevent later launches. Conversely, repeated successful reflights could accelerate activity sharply. NASA campaign timing adds calendar risk: launches delayed beyond December 31 do not count toward 2027. The wide percentile spread preserves both continued development delays and a faster operational transition.
Read the full report
TL;DR
The forecast for January 1–December 31, 2027 (UTC) is a median (50th percentile) of 14.1 launches, a 10th percentile of 6 launches, and a 90th percentile of 30 launches. Plan for substantial growth from development-era activity, not routine weekly operations. The decisive constraint is the supply and turnaround of flight-ready vehicles, not announced pad capacity.
Context
This forecast is dated November 17, 2026. The question counts every worldwide integrated Super Heavy–Starship liftoff during the forecast year, including failed flights, experimental missions and reflights. Static fires, aborted countdowns and standalone-stage flights do not count. The eventual result is a whole-number count; the median above is the saved forecast statistic, not a claim that a fractional launch can occur.
Starship has begun delivering operational payloads and has recovered an upper stage from orbit. Those are material advances. But the latest mission also lost its booster, leaving a gap between demonstrated recovery and dependable full-system reuse (SpaceX Flight 15 report). The forecast reflects that transition: more operational capability, with development risk still shaping the calendar.
Evidence
The strongest historical reference is Starship itself. The complete annual record available at the forecast date is 2 integrated launches in 2023, 4 in 2024, 5 in 2025, and 4 in 2026 through November 17. That makes 15 flights since the first integrated launch on April 20, 2023. This history shows gradual growth, not an established high-frequency operation (launch records).
The recent sequence gives a more useful picture than the annual totals alone. The four 2026 launch dates were May 22, July 24, September 28 and November 13. Their calendar-date gaps were 63, 66 and 46 days. The latest gap shortened, but the sequence is too small to establish a durable trend (launch records). May's flight introduced V3, Raptor 3 and Pad 2, so the low year-to-date count also includes a major vehicle and infrastructure transition (SpaceX Flight 12 report).
This cuts both ways. Extrapolating the entire year would understate the capability developed since May. Extrapolating the latest improvement would overstate the evidence for rapid turnaround. The central forecast therefore requires meaningful operational improvement, rather than either unchanged development cadence or immediate realization of management's ambitions.
Flight 15 supplies the clearest technical reason for expecting growth. It launched on November 13, deployed 26 operational Starlink V3 satellites, and recovered the upper stage on November 14 after approximately 22.5 hours. But the booster ended boostback early because of low fuel mass, diverted offshore and was terminated. The cause remained under investigation (SpaceX postflight report). A successful catch is not yet a demonstrated reuse cycle. Inspection, repair, qualification and reflight still separate recovery from higher launch throughput.
The commercial demand is also tangible. Flight 14 deployed 26 operational satellites, although an upper-stage engine issue shortened its orbital mission (SpaceX Flight 14 report). This matters because Starship need not wait for a lunar campaign to have useful missions. It also shows why payload success and operational reliability should be assessed separately.
SpaceX's October 20 earnings remarks targeted weekly flights at some point in the first half of 2027, with twice-weekly or faster operations later. Management also targeted an upper-stage reflight and orbital refueling attempt early in the year (earnings transcript). These are company projections, not measured throughput. They establish intent and a credible acceleration path, but the historical gaps and incomplete recovery record remain stronger anchors for the central outlook.
Florida offers a route to additional operations. Early-November management remarks described LC-39A construction as effectively complete and targeted vehicle transport in December (reported company remarks). Construction progress is meaningful. Yet transport, acceptance testing, integrated rehearsals and operating authorization are separate gates. A nearly complete pad is not the same as a repeatable launch service.
Vehicle supply is a related constraint. Dated observations identified B23/S43 as a cryoproof-tested pair, while the ship still needed its installed-engine campaign. Other hardware was at earlier assembly or testing stages (hardware observations). I read this as evidence of a developing production pipeline, not a verified stock of flight-ready vehicles. Moving a Texas-built stack to Florida can change its launch location without increasing the worldwide total.
Regulation provides room for growth, but it should not be mistaken for a manifest. Starbase's existing allowance is 25 annual launches. Proposed phases of 100 and 456 annual launches remained at environmental scoping, rather than completed approval (FAA scoping materials). LC-39A's environmental decision covers 44 annual launches, while the FAA explicitly distinguishes environmental review from an operating license (FAA LC-39A page). SLC-37's proposed 76-launch envelope also involves further review and licensing requirements (FAA Cape Canaveral page). These figures are neither expected utilization nor a single worldwide ceiling.
Supporting infrastructure adds another layer of execution risk. NASA's June assessment identified 30 MW of substation capacity but only 9 MW available jointly to LC-39A and LC-39B. It also described aging transformers, potential interruptions of at least one month, and multiyear electrical upgrades costing $136 million (NASA OIG infrastructure assessment). Those are report-vintage findings, not proof of November readiness or a numerical limit on 2027 launches. They explain why towers and launch mounts alone do not establish sustained capacity.
NASA provides additional demand, but no guaranteed tanker total. Its revised architecture places Artemis III's Earth-orbit demonstration in 2027 and lunar landing objectives in 2028 (NASA announcement). The milestone roadmap targets propellant transfer early in 2027 and an uncrewed lunar demonstration later in the year (NASA presentation). These missions depend on technical progress and can move together when a shared prerequisite slips.
The schedule history argues against treating those targets as fixed commitments. Vehicle-to-vehicle propellant transfer moved from March 2025 to March 2026, then to early 2027 (NASA OIG HLS assessment, SpaceX earnings transcript). The lesson is not that the missions will fail. It is that desired demand does not determine which calendar year receives the launches.
What's non-obvious
The obvious story is that a recovered ship plus more pads produces a rapid launch ramp. The missing step is the complete operating cycle. Both stages must become available again, or replacements must arrive quickly enough. A second launch site reduces exposure to local pad downtime, but it does not remove a shared engine problem, vehicle shortage or fleet-wide safety issue. Extra pads can redistribute launches before they increase the total.
The counting rule also changes how reliability matters. A failed flight contributes exactly as much to the immediate count as a successful one. Its damage to the forecast comes afterward, through investigations, lost hardware and repairs. Likewise, full reuse is not required for every useful satellite deployment. The key question is not whether each mission achieves every objective, but whether setbacks interrupt the flow of subsequent flight-ready stacks.
Uncertainties
The largest technical gap is measured turnaround. The evidence does not establish routine V3 booster recovery, repeat upper-stage reuse or a dependable refurbishment interval. Repeated recovery-to-reflight cycles, with documented inspection and repair times, would distinguish a true operational transition from isolated milestones.
Production is another gap. Assembled vehicles, factory work cells and cumulative engine production are not the same as qualified monthly output. Accepted-stage delivery rates, replacement-engine availability and flight-ready inventory would clarify how much growth is possible before reuse becomes routine.
Florida's precise operating-license scope and completion of its full-stack acceptance campaign were not established. That is uncertainty, not proof that authorization or readiness is absent. Verified licensing, completed integrated tests and actual turnaround performance would close much of this gap.
Finally, the lunar campaign's timing and propellant needs remain unsettled. Successful transfer demonstrations would clarify the required work, while NASA's updated mission schedule would clarify when it belongs. A late campaign can shift several intended launches beyond the counting window. These linked uncertainties explain why the forecast preserves both a development-delay downside and a substantial acceleration upside.
Forecast history
| Date | Forecast |
|---|---|
| Median 14.1, 80% range 6 to 30 |
Sources Prolepsia read
- SpaceXwww.spacex.com
- launch recordsll.thespacedevs.com
- Transcript (opens in new window)s21.q4cdn.com
- SpaceX now fundamentally an "infrastructure company," senior official says - Ars Technicaarstechnica.com
- Starship Flight 15, a dual catch attempt, awaits FAA green…www.thespacelaunches.com
- Access the presentation in English herewww.faa.gov
- Final Report - IG-26-010 - NASA's Launch Infrastructureoig.nasa.gov
- NASA, Partners Accelerate Lander Development for 2028 Moon Missions - NASAwww.nasa.gov
- NASA presentationntrs.nasa.gov
- 'Things are going super well at 39A': SpaceX gearing up to launch Starship from Florida for 1st timewww.space.com
The question
This question asks how many launches of SpaceX's Starship launch vehicle will occur worldwide during the 2027 calendar year, from January 1, 2027, through December 31, 2027, inclusive. The forecast is made as of November 17, 2026. A qualifying launch is an actual liftoff of the integrated Starship launch system, consisting of a Super Heavy booster and a Starship upper stage, from a launch pad. All qualifying launches count regardless of mission purpose, payload, destination, flight outcome, or whether previously flown hardware is reused. This includes experimental test flights, orbital and suborbital missions, commercial satellite deployments, propellant-transfer and refueling missions, and missions conducted for NASA or other customers. Launches from any location worldwide count, including Starbase in Texas and potential launch facilities in Florida. Each liftoff counts as one launch, regardless of how many stages or spacecraft are involved. As of November 17, 2026, SpaceX has completed 15 integrated Starship flights since the program's first full-stack launch on April 20, 2023. The most recent, Flight 15, launched on November 13, 2026. Four integrated Starship launches have occurred so far in 2026, following five in 2025. SpaceX is working to increase Starship's launch frequency as the vehicle progresses toward routine operations and full reusability. The Federal Aviation Administration (FAA) authorized up to 25 annual Starship launches from Starbase, Texas, in 2025, and additional environmental reviews are underway concerning higher launch frequencies and operations at other launch sites. NASA's plans for Starship as a lunar landing system also create demand for numerous launches, including orbital refueling missions. The actual number of launches in 2027 will depend on vehicle development, manufacturing capacity, launch infrastructure, regulatory approvals, mission demand, and operational reliability. Relevant background sources include SpaceX's official launch history (https://www.spacex.com/launches/chorus), the FAA's Starship project information (https://www.faa.gov/space/stakeholder_engagement/spacex_starship), and NASA's June 2026 launch infrastructure assessment (https://oig.nasa.gov/wp-content/uploads/2026/06/final-report-ig-26-010-nasas-launch-infrastructure.pdf).
How it resolves
The question resolves to the total number of qualifying Starship launches whose actual liftoff occurs between 00:00:00 UTC on January 1, 2027, and 23:59:59 UTC on December 31, 2027, inclusive. A launch counts when an integrated Starship vehicle, comprising a Super Heavy booster and Starship upper stage, physically lifts off from its launch pad under rocket propulsion with the intention of conducting a flight. A launch counts even if the vehicle subsequently experiences engine failure, loss of control, explosion, flight termination, failure to reach its planned trajectory, or another mission-ending anomaly. Successful orbital insertion, payload deployment, stage separation, booster recovery, and upper-stage recovery are not required. Each distinct liftoff counts exactly once. A launch that is aborted before liftoff does not count, including countdown holds, static-fire tests, ignition attempts that do not result in liftoff, and other ground tests. Standalone flights of Starship upper-stage prototypes without a Super Heavy booster do not count. Standalone Super Heavy booster tests do not count. The primary resolution source will be SpaceX's official launch records, mission updates, and published flight histories. Launch dates and the completion of liftoff will be cross-checked against FAA launch records and other authoritative spaceflight databases, including Jonathan McDowell's General Catalog of Artificial Space Activities and Next Spaceflight. Where records disagree, contemporaneous launch footage, SpaceX mission statements, and official regulatory records will be used to establish whether liftoff occurred and its UTC date. The question will be evaluated after December 31, 2027, allowing up to 60 days, through February 29, 2028, for late reporting and reconciliation of launch records. If SpaceX conducts no qualifying launches during 2027, the outcome is zero. If the Starship program is permanently discontinued, suspended, or delayed during the period, the question still resolves to the number of qualifying launches that actually occurred. If SpaceX's records are unavailable or incomplete, the outcome will be established using the independent sources and contemporaneous evidence described above. The question will be annulled only if the available evidence is insufficient to establish a reliable launch count, an outcome considered highly unlikely.
Fine print
The count is global rather than limited to launches from Starbase, Texas. All Starship variants that use the integrated Super Heavy and Starship architecture qualify, including upgraded versions introduced during 2027, cargo variants, tanker variants, and Human Landing System variants, provided they launch atop a Super Heavy booster. A Starship spacecraft launched into space and subsequently refueled, moved between orbits, landed, or relaunched from a celestial body does not generate additional qualifying launches unless it undergoes a new integrated Super Heavy–Starship liftoff from a launch pad. Reflights of previously recovered Starship hardware count as separate launches whenever a new qualifying liftoff occurs. The calendar year is determined by Coordinated Universal Time (UTC), not local time at the launch facility. Launches occurring before January 1, 2027, or on or after January 1, 2028, UTC are excluded even if they are associated with missions scheduled for 2027. The numerical forecast range of 0 to 60 launches is intended to accommodate both continued low-frequency development operations and a substantial increase in launch cadence. It is a forecasting range, not a regulatory limit or a claim about SpaceX's planned launch frequency. Outcomes above 60 remain theoretically possible. Although launch counts are necessarily nonnegative integers, the numeric question type is retained to match the requested forecasting format and draft values.
More public forecasts
Will less than 5 SpaceX Starship launches successfully reach Space in 2026?
85%chance of yes
Prolepsia forecast
Will SpaceX's Starship deploy operational Starlink satellites into orbit on at least 5 separate flights between 1 January and 31 December 2027?
76%chance of yes
Prolepsia forecast
Will SpaceX launch at least one of its AI1 data-centre satellites before 1 January 2028?
60%chance of yes
Prolepsia forecast
Will SpaceX have the highest IPO Market Cap 2026?
98%chance of yes
Prolepsia forecast