This look at the Jeep prototypes developed by Willys-Overland engineers between 1949 and 1953, leading up to the CJ-3B, M38A1 and CJ-5 models, incorporates information from research by Jim Allen, Bill Munro, Ron Szymanski, Todd Paisley, Bill Norris and Bob Westerman. The drawings are by Motonobu Sato, and are courtesy of J4 magazine. (Note: See a complete set of the drawings in a new window for easier comparison, 85K GIF.)
Because many of these prototypes were being developed simultaneously, the timeline can be confusing. To help clarify the order in which they appeared, see the chronological list of significant dates in the Jeep Prototypes Timeline 1948-53.
Footnote numbers in the text(1) refer to the List of References at the bottom of this page.
The CJ-3A was the civilian Universal Jeep in production by Willys-Overland during this period (see the CJ-3A Information Page.) It evolved from the CJ-2A in 1948, retaining the L-head 4-cylinder engine, as did the M38 military version built during 1950-52. A thousand CJ-3As were also produced in modified form for the U.S. Navy as the submersible CJ-V35 in 1950. (4)
This photo was taken from the Nebraska Tractor Test of the CJ-3A in October 1949. The 3A was manufactured until shortly after the debut of the 3B in 1953, but as early as 1949 Willys was working on designs to replace the 3A, with a Jeep powered by the slightly larger F-head "Hurricane" engine.
On 29 June 1949, Engineering Release 5607 was opened by the Willys Engineering Dept., with the title "Front End Redesign ('FB' Head Engine Necessitates New Design) (Model CJ-3A)." (6) Although we have not seen factory photos to confirm what the first design looked like, I have long assumed that it was similar to the final CJ-3B design.
This may have been confirmed by the discovery of the Jeep in this photo taken in 2017, with serial number CJ3B-01 stamped on a Willys-Overland CJ-3A patent plate of the type used from early 1949 to mid 1950. The most visible difference from a later CJ-3B is the air vent on each side of the cowl. See more details about CJ3B-01 on CJ3B.info.
This was the original modification referred to by Jim Allen as the "hoodectomy," to make room for the F-head. Other than just cutting a hole in the hood to clear the F-head's carburetor (as O.E. Szekely Inc. did to put the Hurricane Engine in a Low-Hood Jeep when they built aircraft starter Jeeps for the Navy and Air Force,) simply raising the cowl, grille and flat hood would seem like the most obvious solution.
It also seems logical that the engineers might have tried a sloping hood solution, similar to the 1953 photos seen further down this page.
The earliest prototype photo we do have is from August 1950, shortly after Willys was finally granted United States Trademark Registration number 526175 for the name "Jeep," on 13 June 1950. A Jeep bearing the experimental vehicle number X-98 has a raised cowl, and a new hood stamped with a slight curve down in the front and a raised panel in the center. X-98 retained the CJ-3A's flat fenders, but with a front clip not unlike the eventual CJ-5 grille.
Jim Allen has written: "I have several pics of this rig taken at different times and in different markings. In a set of Army photos taken at the Detroit Arsenal and dated 24 August 1950, this vehicle bears the number X98 on the bumper and is called the 'truck, 1/4 ton, 4x4 with F-head engine.'"(10) See also the left side and front view (40K JPEGs).
This Jeep was a civilian prototype, but the Korean War had begun in June 1950, and the Army was anxious to get a more powerful Jeep to deal with the mountainous terrain there.
Project 5607 was initially closed on 3 August 1950. But what appears to be the same prototype was photographed being further tested by the U.S. Army the following year at Fort Knox KY, now with the number "205" on its front bumper.
Perhaps 205 (formerly known as X-98) was the first complete prototype to emerge from Project 5607, although the new Project 5707 was already well underway. Engineering Release 5707, "New Model Release 473CJ" had been opened on 21 March 1950.
Jim Allen says, "A test done 5 April 1951 at Fort Knox, included two WWII-era MBs, two M38 model MCs, a standard commercial CJ-3A and an F-head rig they referred to as the "F-head truck." Each vehicle was given a test number, the F-head truck's being '205'. I believe X-98 is the vehicle used for those '51 tests at Fort Knox. Fred Crismon told me that this picture on page 224 of his book (2) was dated 8 May 1951." (CJML (5) 12 Oct 1999.)
The marking on the hood refers to "Army Field Forces Board No. 2", the Armored Board at Ft. Knox. (13)
This photo of 205, a left side view (270K JPEG) and a front view (130K JPEG) dated 1 August 1951 have captions reading "Project 5707, Model CJ-4M (Test vehicle loaned to Fort Knox)." (12) This caption connects the Jeep to Engineering Release 5707, but the reference to CJ-4M is misleading. Jim Allen has speculated that when a military version of the CJ-4 was being contemplated (below), the old X-98 was brought in as a stand-in, and given the designation "CJ-4M" until the real thing was built.
The CJ-4M had in fact been photographed complete at the factory in March 1951, but it does appear that "205" was mechanically similar to the CJ-4M and was considered suitable for testing.
This is where things start to get confusing, with the same prototype being tested or photographed more than once, with different model numbers or markings. One author, in J4, states that "The civilian version of CJ-4M existed as X-98 in 1950." (1) X-98 did have civilian features such as a tailgate, side-mounted spare, and "WILLYS" on the hood, although it was also tested by the military. However, X-98 was clearly a different design than the CJ-4 and CJ-4M (below).
See Bob Westerman's replica of X98 in Legendary X98 Brought to Life on CJ3B.info.
X-98 may have been a result of either Project 5607 or 5707, but the "CJ-4" design was certainly part of Project 5707, which continued into the middle of 1951. This next group of prototypes had different front-end bodywork, including distinctive front fenders which were skirted all around, and a unique front grille.
Todd Paisley notes about the paperwork for Engineering Release 5707: "Whatis interesting is the model number given under the model number field. Itfirst was a '473 CJ'. That was crossed out and '474 CJ' was entered. Thenthat was crossed out and a 'CJ-4' was entered.
"Willys at the time was condensing all their model lines to a single number. The letter combinations after the 473 would designate the type of vehicle. (i.e. SW for Station Wagon, 4WD (or HT) for truck, VJ for Jeepster, SD for Sedan Delivery, PD for Panel Delivery, etc.) It looks like they originally were going to increment the number each year because a year later was a bunch of entries for a 474 model series that didn't make it to production. (Station Wagon was one of them. Project 6370 was "New Model Release - 474SW-SD-DSW". All existing models except the CJ had a 474 new model release.) Project 5707 outlasted this, so the 473CJ entry became the 474CJ entry. When the 474 designation never materialized, it was crossed out to CJ-4." (CJML (5) 20 Nov 1999)
A CJ-4 is seen with a 1950 Ohio license plate in this photo from page 77 of the Arch Brown book, where it carries the number X151 on its bumper. (3) Originally painted gloss green, this is apparently the same Jeep which has survived in reasonably good condition, repainted in red (see Discovering the Lost CJ-4 by Jim Allen on CJ3B.info.) Mechanically it is similar to the subsequent MD model (M38A1), but with standard rather than reversed front spring shackles.(7)
Engineering Release 5707 was closed on 6 June 1951.
The first evidence of a military version of the CJ-4 was a heavily retouched illustration (90K JPEG) which may have been a doctored photo of the civilian version.
Opened on 28 December 1950, Engineering Release 6555, "Military Jeep with 4FB Engine (Test & Design one Military Jeep) (Model CJ-4M)," was the project leading to the military version of the CJ-4, shown here in factory photos dated 5 March 1951. (12) This was a fully military vehicle, with snorkel, blackout lights and external electrical connectors, but no photos have yet surfaced showing it being tested by the Army, which was still testing the earlier X-98 prototype during the spring and summer of 1951.
Project 6555 was closed on 21 June 1951. See more details on The Military CJ-4M on CJ3B.info.
Note: The CJ-4M has been referred to as the M38E1, but photographic evidence suggests that the U.S. Army designation M38E1 actually refers to vehicles based on the M38. Research by Bob Westerman has produced Aberdeen Proving Ground photos showing an XM38E1 (210K JPEG) on 23 August 1951 and an M38E2 (210K JPEG) on 12 December 1951. Bob says these new versions were produced for a military "Austerity Program," to reduce the cost of the M38.
The fact that Willys was still offering the Army versions of the M38, while working on the new F-head prototypes, is also illustrated by Engineering Release 6396 "4-Wheel Drive Ambulance -- Research Tests," opened on 5 October 1950. It listed the model as the "MC-A" which would suggest an ambulance version of the Army's M38, designated "MC" by Willys. A 21 January 1951 factory photo of an MC-A ambulance (50K JPEG), is an early result of the 6396 project. Another photo (30K JPEG) appeared in Kaiser Corporation's 1953 Annual Report, illustrating a mention of the continuing development of military Jeeps.
But once again it appears that work on more than one prototype may have been included under a single Engineering project. Factory photos dated 23 March 1951 are labelled "Project 6396, Model CJ-4MA-01 Ambulance" and show a long-wheelbase military prototype based on the CJ-4 design, captioned variously as ambulance and personnel carrier (60K JPEGs). (12) This prototype was found to still exist in 2004; see CJ-4MA Military Ambulance Discovered.
In Army testing (seen at left in May 1951) the CJ-4MA was also known as XM170, and eventually evolved into the M170 ambulance.(11)
Engineering Release 6825, "Front Line Personnel Carrier 100" W.B. 4 Cyl. F-head (1500# Payload Plus 2 Pass.) (model CJ-4MP)" was opened in April 1951, apparently a variation of the CJ-4MA ambulance from Project 6396. No photos are known to exist. There is however an undated photo of a version of the CJ-4MA with dual wheels (40K JPEG).
All of these Engineering Releases involved much more than body design concepts. Surviving Willys Engineering Calculations show detailed figures for spring design, maximum grade and drawbar pull for the LWB military prototypes.
6 October, 1950, Engineering Release 6400 "Jeep-Advanced (1/4-ton 4x4, AJ)" was opened, and cancelled 15 March 1951. Nothing is known about this experiment. Bill Munro suggests it was an attempt to begin development of an all-new military vehicle, and was scrapped when Ford was the only company invited to tender for what eventually became the M151 MUTT.(11)
Meanwhile, on 22 January, 1951 work began on another military Jeep described as "advanced," but commissioned by the Army and based on the F-head engine.(11) Engineering Release 6600 was titled "Advanced Military Jeep (New Model Rel.) (Model MD)."
Project 6600 closed 28 March 1952, and by June the first MD, or M38A1 as it was known by the Army, came off the assembly line. Bigger, heavier and stronger than earlier Jeeps, the M38A1 also was the first appearance of the rounded front fenders. Over 100,000 were built from 1952-1957, and some lasted into the 1970's in U.S. military service.(9) The subsequent CJ-5 and its close relatives were based on this design.
Willys-Overland president Ward Canaday had announced in May 1951 that Willys and Ford (Canada) would study possible production of a quarter-ton 4x4 in Canada, and in October 1951 that Willys would provide Ford with plans and technical assistance to build Jeeps in Canada.(14) The model was not specified, but initially was the M38, and Ford built 2,400 of the M38 CDN. Some 1,000 A1's were also built by Ford in Canada in 1953 and designated M38A1 CDN, for delivery to the Canadian military. (See Willys Model MD on CJ3B.info.)
On 30 November, 1951 Engineering Release 7474, "4 FB Jeep (Model CJ-4A)" was opened, and closed only a month later, on 3 January 1952. This project briefly introduced the model designation CJ-4A, but if there was in fact a different prototype built, no photographs are known to exist.
The CJ-4, like X-98, did not get beyond the testing phase. Project 5607 was reopened on 12 January 1952, further changes were possibly made, and Willys finally gave the world the CJ-3B, which seemed like a bit of a throwback.
This photograph of the result of Project 5607 confirms that it was basically the production CJ-3B. Bill Norris found a copy of the photo at the Detroit Public Library, with the original Willys-Overland typewritten caption: "Project 5607, Model CJ-3B Jeep, 3/4 Right Rear View (Top On -- Complete), Engineering No. 13410-15, 11-3-52." The full 7-piece Willys-Overland Canvas Top, modified to fit the CJ-3B, was apparently still part of Project 5607.
By 3 November 1952 when this photo was taken, the CJ-3B was already in production. Engineering Release 5607 was finally closed on 6 April 1953, four years after it was originally started.
The question of why the CJ-4 and its variants never went into production has still not been fully answered, but in 1999 Todd Paisley suggested the following, based on his research: "The Engineering Releases during this time frame had lots of changes to reduce the amount of critical materials because of the Korean War. Entries made it clear that chromium, aluminum and brass were considered critical materials, and efforts were made to replace them with other materials.
"What probably happened with the CJ-4 was that adequate tooling could not be procured to produce the model (since it didn't relate to the war effort), so the effort to modify a CJ-3A was looked at again. I think it was a matter of not having the resources to produce the model, rather than some inherent problem with the design." (5)
Even after work began on the model MD (M38A1) in 1951, Willys was restricted in its ability to turn this into a civilian model because the Army had commissioned the MD and owned the design. Research by Bill Munro suggests it took some time for Willys to negotiate for the right to build the CJ-5 (see below).
The fact that Willys had also decided in 1949 to develop a new passenger car, the Aero Willys, would clearly have had an impact on Jeep development. The relatively small company didn't have the same cash resources as the Big Three automakers, to finance new models.
I'm also going to speculate that the need for a simple design which could be built in other countries may have been a contributing factor. The CJ-3B was exported to many countries around the world (see CJ-3Bs for the World) and the local manufacturing of parts (and eventually complete vehicles) was what allowed the Jeep to be so successful in the postwar era of prohibitive tariffs on automobile imports, designed to stimluate local industry. It made the high-hood CJ-3B by far the most common Jeep design around the world. The two largest producers, Mahindra and Mitsubishi, later added dog ears to the front fenders, but it would be many years before Mahindra produced rounded hoods and fenders, and in the case of Mitsubishi, never.
This photo is labelled "CJ-3B Hood Mock-Up -- Left Front, Willys-Overland Engineering No.13195(1)" and is dated 6 June 1952. Since the vehicle is labelled as a CJ-3B, this more sophisticated hood design may still have been part of Project 5607, which was not closed until 6 April 1953. A front view photo (70K JPEG) shows that two proposed stampings are grafted together in the center.
This straight but sloping hood is possibly another experiment from Project 5607, this time produced under instruction to keep the stamping simple. It seems to have a fully-developed CJ-3B cowl and ventilating windshield, and it appears to be the same prototype seen in the May 1953 outdoor photos below. See also a front view photo (50K JPEG).
Jan Hogendoorn first drew my attention to this photo of the sloping hood, which had been published in the August 1986 issue of Collectible Automobile magazine, accompanying an article called "Before Wrangler: CJs That Never Were." Author Jay Koblenz wrote, "Willys tried grafting a downsloped hood onto a CJ-3B in 1953, with the aim of improving forward visibility. The idea never made production, but a surviving photo from Willys Engineering suggests that it probably would have improved aerodynamics as well...."
I now have an original print of the photo, complete with caption confirming Koblenz' description (see the large copy, 120K JPEG). The date given is 5 May 1953 which could place it as the final proposal from Project 5607.
Rear view photo 14104-3 has the same date in its caption. It does seem strange that Willys Engineering would have gone back to experimenting with the CJ-3A radiator guard, after all this time.
(Note: A similar hood was used on the Lycoming JB-1 Aircraft Starter. A Jeep seen for sale on the web (40K JPEG) in 2009 was also a similar conversion.)
The CJ-3B was already in production when these new hood ideas were being tried, so it appears likely that if one of them had been approved, it might have become a 1954 model. Would that have been the early end of the 3B as we know it? Would it have been given a new model number? It's all speculation now anyway; the go-ahead for the CJ-5 would put an end to these experiments.
Meanwhile, the Engineering Department was also considering changes under the hood. Engineering Release 9327, "CJ-3B 12 VoltExperimental Pilot Model (Japanese Police Reserve) (For JapaneseGovernment)" was opened on 30 January 1953. Note that production CJ-3Bs at this date were all 6-volt. And on 10 September 1953 it opened Engineering Release 9811, "Special CJ-3B With Deutz 4 Cyl. Air Cooled Diesel Engine (To Study 4 Cyl. Air Cooled Diesel Engine)."
The M38A1 went into production in June 1951, so why didn't Willys release a civilian round-fendered Jeep instead of the CJ-3B in 1952? The CJ-5, the civilian Jeep derived from the MD design, did not enter production until the 1955 model year, allowing two years of glory for the CJ-3B as the only Universal Jeep.
The J4 author speculated, "Willys wanted to sell the CJ-5 as M38A1s civilian version, but in 1953 all production lines were occupied by the M38A1, no space for the CJ-5. But on the CJ-3A production line, it was possible to produce a new Jeep converted from the 3A -- the CJ-3B." (1) This explanation is not completely convincing, since Willys had in the past produced multiple models concurrently on one production line, and would continue to do so (see Working at the Overland).
The pre-production CJ-5 seen in this undated photo found by Justin Kennedy has a "Willys" rather than "Jeep" stamping on the cowl, and has a split windshield. Other photos of this vehicle appear in a Willys Product Merchandising Manual (170K JPEG) aimed at dealers, circa 1954. But a much earlier illustration which appeared on the October 1952 cover of Motor Trend (330K JPEG) also looks like it may be based on this civilian prototype.
Is it possible this civilian version of the M38A1 was the subject of Engineering Release 6400 (the "AJ") in late 1950, or Engineering Release 7474 "4 FB Jeep (Model CJ-4A)" in late 1951? We have no factory photos or documentation from either of these projects.
Research by Bill Munro suggests that the reason Willys took so long to release the CJ-5, is that since the Army had commissioned the MD they owned the design. Willys had to secure exclusive rights to the civilian version from the government; this is a plausible explanation for the mysterious delay. Munro states, "Willys did not have enough money to build a completely new body for a new Hurricane-powered CJ. A compromise had to be found, and that was the CJ-3B."(11)
As I said at the top of the page, many of these prototypes were being developed simultaneously, so the timeline can be confusing. To help clarify the order in which they appeared, see the chronological list of significant dates in the Jeep Prototypes Timeline 1948-53.
Further information on the prototypes is always welcome. Perhaps the most tantalizing missing photo is factory evidence of "CJ3B-01" under Engineering Release 5607 in 1949.
Thanks to all the authors referenced above, and to Bill Norris and Bob Westerman. Thanks to Charlie Weaver and the Willys-Overland-Knight Registry Inc., guardians of the Engineering Releases. -- Derek Redmond
Also on CJ3B.info, see a humorous look at the origins of the high hood in The Jeep CJ-3B Revisited, and some other Jeep design concepts in Jeep Designers at Work.
Return to Universal Jeep History.
Visit CJ3B.info on Facebook.
CJ3B Home | Contents | Search | Links | 3A and 3B Community