The Titan II/III/IV family (derived from the ICBM family of the same name)1 were the premier U.S. expendable launchers for much of the late 20th century, being the most capable expendables in the U.S. inventory from 1973 (with the retirement of the Saturn V) until 2002 (when the Atlas V and Delta IV entered service). They were also some of the most-expensive-per-unit-capability launchers ever put into service, causing them to be used almost exclusively for military satellites, along with twelve mostly-manned Gemini flights and a handful of experimental and science satellites; due to the launchers’ high cost-density, only three commercial satellites were ever launched by Titans, all of them by the unsuccessful Commercial Titan III variant in January-June 1990.2 Even the military, with radically-different launcher-selection priorities from commercial customers, found the high cost of Titan launches to be a serious hindrance, leading to the Titan family’s eventual retirement in favour of the newly-developed Atlas V and Delta IV.

I’m at somewhat of a loss as to why the Titans’ cost-per-capability was so high. One seemingly-obvious explanation would be the considerable handling difficulties resulting from their use of highly toxic and explosive hypergolic propellants (requiring extensive protective equipment and garb for the workers)... except that the Russian Proton and Chinese Long March families both had long, profitable careers as commercial launchers (indicating that their cost-to-capability ratios were acceptably low), despite using hypergolic propellants themselves, and, thus, requiring at least Titanesque degrees of expenses to mitigate the hypergols’ handling difficulties (probably considerably greater, in fact, since launches out of Baikonur or Xichang or Jiuquan or Taiyuan drop spent stages over land, incurring large cleanup costs when the incoming stages are hypergolically-fuelled, while launches out of Cape Kennedy or Vandenberg drop their stages harmlessly3 in the ocean).

So why were the Titans such expensive rides?

1: The Titan I was used solely as an ICBM; the Titan II pulled double duty as an ICBM and a space launcher; and the Titan IIIA, Titan IIIB, Titan IIIC, Titan IIID, Titan IIIE, Titan 34D, Commercial Titan III, and Titan IV served as dedicated space launchers.

2: The Commercial Titan III carried a total of five payloads in its four launches, but one of the first two (the payloads on the first CT-III launch, just after midnight on New Year’s Day 1990) was a British military satellite, and the last (launched in September 1992) was the NASA Mars Observer probe.

3: Harmlessly to people, that is; probably not harmlessly to the local marine life, but said marine life generally lacks the ability to lobby the government or sue the launch provider to get them to stop dropping spent rocket stages in the ocean.

  • 2
    $\begingroup$ Do you have references for the launch costs of Titans and their competitors? $\endgroup$ Jun 23, 2020 at 1:39
  • 1
    $\begingroup$ I'm placing a bounty because my guess of an answer is insufficient & I woud love to know the real reason $\endgroup$ Sep 8, 2021 at 23:06
  • 1
    $\begingroup$ @RussellBorogove I'm not seeing specific sources for the launch costs, but Astronautix has them listed. TII: Recurring Price $: 16.389 million in 1969 dollars (astronautix.com/t/titanii.html). TIII: Launch Price $: 136.600 million in 1992 dollars. (astronautix.com/c/commercialtitan3.html). TIV: Launch Price $: 400.000 million in 1985 dollars. (astronautix.com/t/titan4.html). $\endgroup$
    – paulmrest
    Sep 9, 2021 at 0:04
  • 3
    $\begingroup$ In 2021 dollars: TII, \$122M (\$40M/ton-to-LEO); TIII, \$264M (\$18M/ton); TIV, \$1015M (which sounds suspect; Astronautix is not a great source). I found a source saying \$253M in 1995 for a Titan IV, so \$453M (\$21M/ton). (Pro tip: search for (rocket) cost GAO hearing to find the numbers the manufacturers didn't advertise, but government had to discuss. $\endgroup$ Sep 9, 2021 at 2:43
  • 1
    $\begingroup$ The safety standards that the different countries are operating under are different though. Bluntly put, neither China nor Russia actually care about the environmental or human cost of unsafe materials. If a spent upper stage crushes some peasants or launch complex workers all develop cancer... well, that's just the price of progress. Sure, the USA wasn't (and isn't) a paragon of worker's rights and safety procedures, but compared to China and Russia? $\endgroup$
    – Dragongeek
    Sep 10, 2021 at 11:34

2 Answers 2


This is not an answer, but some relevant data, and a wild guess.

It's difficult to find apples-to-apples comparison of launcher costs, but I found a terrific source: a 1990 General Accounting Office fact sheet titled Military Space Programs: An Unclassified Overview of Defense Satellite Programs and Launch Activities. Among other interesting contents, this has an appendix detailing the various launchers contracted by the Department of Defense at that time, with payloads to 100nm altitude Easterly LEOs given, and the DOD's per-launch costs (apparently actual negotiated, contracted costs provided where possible, and estimates for the rest).

So with this nicely normalized data, it's very easy to work out the cost per LEO-payload-mass of the various launchers used by DOD in the early '90s (dollars US 1990, tons metric):

  • Titan IV (no upper stage): \$8M / ton-to-LEO

  • Titan III: \$9.5M / ton

  • Titan II: \$22.6M / ton

  • Shuttle (DOD cost)1: \$5.4M / ton

  • Atlas II: \$6.8M / ton

  • Delta II: \$7.3M / ton

  • Delta I (4000-series?): \$9.6M / ton

  • Shuttle (realistic cost)1: ~\$12M / ton

  • Atlas-Centaur2: \$16.8M / ton

  • Atlas E: \$57M / ton

note 1: The cost given for the shuttle in the GAO document is impossibly low; I think this is because ongoing training and facilities expenditures were coming out of NASA's budget, not DODs -- it's notoriously difficult to determine the "actual" cost of shuttle flights. I added a "realistic" cost estimate based on another source.

note 2: The GAO doc doesn't give LEO payload or cost for Atlas-Centaur; I worked those out from a different source.

So, the crazy outliers are the Titan II and Atlas -- the oldest and smallest launchers, both entirely conceived as ICBMs, thus not designed to be cheap commercial launchers, and not optimized for big LEO payloads -- Atlas E is a stage-and-a-half launcher, barely capable of making orbit at all.

Ignoring the 1950s ICBMs, the other Titans and the shuttle (a pineapple in our apples-to-apples comparison) for the moment, we don't have a lot of data points, but they nod towards general cost reduction per payload ton in later designs.

If the higher cost of the late model Titans (\$8-9.5M) relative to the Delta II and Atlas II (\$6.8-7.3M) has an explanation, my guess it's just that there was much more of the very old Titan II design heritage left in the Titan III and IV. The other launchers invested more money and/or effort in redesign, driving per launch cost down.

  • 1
    $\begingroup$ Maybe my "Titan II conversation" answer wasn't so off... just for conversion to launch vehicles; nothing to do with Titan IIIs. $\endgroup$ Sep 9, 2021 at 15:43

Here is my unsourced "common-knowledge" answer. Hopefully it sets somebody more industrious on the right track to find a sourced answer.

Titan II (built by Martin) went into production as a liquid powered ICBM. Once it began to be replaced by solid-fueled Minutemen, the contract was won by Martin Marietta in 1986 to convert Titan IIs into OLVs.

Converting Titan IIs into Titan IIIs was no easy process. Besides extending all four tanks of the two Titan II stages, (for Titan IIICs and later) the Titan II tank was not intended to take the nonaxial stresses of those big UA1205s on the side. Like Boeing is rediscovering wrt. the ET vs the SLS main tank, the differences between tanks taking side-loaded and top-loaded stresses are substantial.

The cost of converting Titan II tanks into Titan III tanks was an expensive process that could not be tooled, giving Titan the reputation of "being each built by hand".

Again, unsourced, but it's what I've always known/heard from other space history enthusiasts.

  • 3
    $\begingroup$ The Titan IIIs weren't converted IIs; they were all built as various iterations of Titan III. The surplus IIs were converted to the 23G configuration to launch satellites, but stayed IIs. $\endgroup$
    – Vikki
    Jun 24, 2020 at 20:50
  • 1
    $\begingroup$ My unsourced common-knowledge answer might be horribly wrong. If you have a source proving me so I'll delete this in shame... $\endgroup$ Jun 24, 2020 at 22:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.