JB23 Automatic vs Manual: Which Holds Value Better in Japan

Published 2026-08-20 ・ 20 min read ・ Japan Vehicle Data

The JB23 was sold in Japan as both an automatic (4AT) and a manual (5MT) throughout its entire twenty-year production run — 1型 in 1998 through 10型 in 2018. If you are buying one for export, both options are in the market right now. The question is whether the transmission changes the price enough to matter.

The short answer: in the retail market, AT and MT prices overlap heavily at any given mileage and year. But in the auction/wholesale market and in the older ends of the range, MT has been documented to command a stronger premium as cars age — particularly for high-type, low-mileage cars where enthusiast demand concentrates.

This article presents market data collected 2026-08-20, explains what drives the difference, and draws out what it means for import buyers. For price ranges by revision type, see JB23 which year to buy. For full mechanical differences across types, see JB23 revision types.

1. Why transmission matters for hold-value in the JB23 context

In most kei-car segments, automatic transmission commands a retail premium because the customer base skews toward convenience buyers. The JB23 is an exception — and understanding why explains the price data.

The JB23 attracted three distinct buyer types: daily-use commuters (who preferred AT), weekend off-roaders (who split roughly evenly), and the growing group of customisers and enthusiasts building lifted, engine-tuned machines for export markets and domestic motorsport. The last group concentrates heavily on MT. This demand structure means the AT’s convenience premium is partly offset by the MT’s enthusiast premium, and the balance shifts depending on what segment of the market you are looking at.

The other factor: the JB23’s enthusiast demand has grown since the JB64 launched in 2018. The JB64 is harder to source and commands a larger price, making late JB23s — especially MT variants in strong condition — an alternative for buyers who want a tunable, exportable platform. US demand for early types (1型–3型, which became 25-year eligible from 2023 onward) has also created a floor on prices in the US-eligible range that would otherwise have continued to decline with age.

2. One correction to common guidance

The JB23 was not MT-only from any type onwards. Automatic transmission was available in every revision from 1型 (1998) through 10型 (2018). At the 5型 revision (January 2002), the AT shifter changed to gate-type operation (“ゲート式”) — but the automatic itself was retained throughout the production run.

This matters to buyers because claims about “AT discontinued after type 4” will incorrectly narrow your search. If you are looking at a 2017 JB23, it can be either AT or MT.

3. Price data: retail listings

How this data was collected

We conducted four separate searches on Japanese retail listing platforms and recorded all visible listings. Collection date: 2026-08-20.

• CarSensor “ジムニー JB23 AT” freeword search: 47 total listings; 19 individual prices captured from page 1 (1 duplicate removed)
• CarSensor “ジムニー JB23 MT” freeword search: 32 total listings; 27 individual prices captured (2 pre-JB23 entries excluded)
• Goo-net JB23 AT filtered search (mission=AT, phrase=jb23): 13 listings; all 13 prices captured
• Goo-net JB23 MT filtered search (mission=MT, phrase=jb23): 16 listings; all 16 prices captured

These are retail asking prices, not auction prices. The Goo-net AT/MT filtered searches use keyword matching against listing descriptions, so some results may cross-contaminate; we have flagged anomalies below. The CarSensor freeword searches may include JB33/JB43 and other generation Jimnys in small numbers — where year/price combinations suggest a non-JB23, we note this.

Total observations: N=76 retail listings (47 AT + 13 AT from Goo-net duplicates excluded; N=47 AT CarSensor + N=13 Goo-net AT, and N=32 MT CarSensor + N=16 Goo-net MT). Because both platforms likely list some of the same cars, the true unique-car N is lower, and we are not computing a merged average. We use each platform separately for directional comparison.

CarSensor AT listings (N=47 total, 19 sampled)

Price range: ¥35.8万–¥201.7万

Sample from page 1 (19 listings captured; one duplicate entry removed):

Note: Two listings (1997, ¥125万 and ¥109万, 3AT) were excluded from the table above as they appear to be JB33/JB43 generation vehicles based on year and 3AT designation. A 3-speed automatic was not fitted to JB23.

Source: https://www.carsensor.net/usedcar/freeword/%E3%82%B8%E3%83%A0%E3%83%8B%E3%83%BC+JB23+AT/index.html — retrieved 2026-08-20

CarSensor MT listings (N=32 total, 27 sampled)

Price range: ¥38万–¥199万

Sample (27 listings retained; 2 pre-JB23 entries excluded — see note):

Entries at ¥125万 (1994) and ¥189.8万 (1995) excluded — these pre-date JB23 production (which began 1998) and are likely JB11/JB33 vehicles.

Source: https://www.carsensor.net/usedcar/freeword/%E3%82%B8%E3%83%A0%E3%83%8B%E3%83%BC+JB23+MT/index.html — retrieved 2026-08-20

Goo-net AT listings (N=13, all captured)

Price range: ¥39.9万–¥179万

Source: https://www.goo-net.com/cgi-bin/fsearch/goo_used_search.cgi?category=USDN&mission=AT&phrase=%E3%82%B8%E3%83%A0%E3%83%8B%E3%83%BC+jb23&query=%E3%82%B8%E3%83%A0%E3%83%8B%E3%83%BC+jb23 — retrieved 2026-08-20

Goo-net MT listings (N=16, all captured)

Price range: ¥63万–¥208.2万

Note: One 1998 listing at ¥208.2万 with 18.6万km stands out — at those prices a 1998 car is likely a heavily restored or collector example, or miskeyed. We include it but flag it as an outlier. The 1997 listing is pre-JB23 and excluded from price analysis.

Source: https://www.goo-net.com/cgi-bin/fsearch/goo_used_search.cgi?category=USDN&mission=MT&phrase=jb23&query=jb23 — retrieved 2026-08-20

What the retail data shows

Comparing same-era cars across the two datasets:

2017 model year — direct comparison:

• AT: ¥159万 (3.4万km), ¥201.7万 (2.7万km) — CarSensor; ¥179万 (5.3万km) — Goo-net
• MT: ¥170万 (5.7万km), ¥170.3万 (4.8万km), ¥147万 (6.5万km) — CarSensor; ¥170.3万 (4.8万km), ¥165万 (4.3万km) — Goo-net

Observation: At comparable mileage, 2017 MT and AT prices overlap substantially in retail listings. The highest AT observed (¥201.7万, 2.7万km) exceeds the highest MT observed, but this may reflect a specific grade or dealer premium. Neither transmission clearly dominates at the retail level in this year band.

2013 model year:

• AT: ¥139.8万 (6.3万km), ¥117.7万 (4.7万km CarSensor) — Goo-net; ¥83.4万 (9.7万km)
• MT: ¥170万 (5.9万km), ¥109.8万 (9.8万km), ¥99.8万 (10.4万km)

Observation: The ¥170万 MT at 5.9万km is materially above comparable-mileage AT examples at ¥117–140万. Higher-priced MT examples at lower mileage appear in both platforms.

Early types (2001–2007):

• AT range: ¥35.8万–¥88万 across 2001–2007 CarSensor entries
• MT range: ¥39万–¥93万 across 2001–2007 CarSensor entries

At early type ranges, MT and AT are roughly equivalent in the retail market, with MT slightly higher at comparable mileage.

4. Auction data: where the hold-value signal is clearer

Retail asking prices are set by dealers and reflect their margin assumptions. Auction prices — where the wholesale market actually clears — are a better signal of structural value.

Source: Used Car Market Log, July 2026 auction settlement data (USS network)

The site tracks JB23 auction prices by year and transmission. July 2026 data shows:

Source: https://usedcar-market-log.com/market-price-jimny/ — July 2026. Evaluation 5 = clean condition, no repair history.

For the 2017 model year — the only year with both AT and MT auction data in this source — AT commanded ¥123万 against MT’s ¥99万, a ¥24万 premium for automatic.

This is the opposite of what some retail-market impressions suggest. The site’s commentary notes that “ランドベンチャーAT車は全年式で軟化” (Land Venture AT cars are softening across all years), while MT is “まちまちで方向感なし” (mixed, no clear direction). This suggests AT held higher in the auction market for this cycle, but has been softening, while MT has been volatile.

The same source notes that for the JB64 (current generation), AT commands a ¥15–28万 premium over MT — a much clearer gap. For JB23, the AT premium in the auction market exists in the data available but is narrower and less consistent.

5. Why the picture is complicated: different buyer segments

The JB23 AT vs MT hold-value question does not have a single answer because different segments of buyers value the two transmissions differently.

The convenience/utility segment drives AT demand among domestic Japanese buyers — daily commuters, older buyers, women buyers. This is the same segment that drives AT premiums across most kei car categories.

The enthusiast/export segment has a different calculus. International buyers — particularly in markets where right-hand-drive SUVs with manual transmissions are valued — tend to favour MT JB23s. Japanese professional off-roading culture also leans MT. The customiser community overwhelmingly works with MT cars (the tunable 5MT drivetrain is the starting point for most builds). This demand puts upward pressure on MT pricing in specialist and export channels.

The buy-back market (dealer repurchase) shows a longer-term pattern: one major Japanese buy-back service states that “年式が古くなればなるほどMT車の査定額がAT車を大きく引き離して高額になる” — as cars age, MT valuations increasingly separate from and exceed AT. This is a qualitative statement, not a price series, but it reflects observed dealer experience with older JB23 cohorts.

Source: https://www.hanamaru870.net/column/jimny_satei/ — retrieved 2026-08-20

The synthesis: In the short term (recent-year cars, 2015–2018 range), AT appears to command a modest premium in the wholesale auction market. In the medium-to-long term, as cars age and the ownership pool shifts toward enthusiasts and niche buyers, MT appears to hold value better. This pattern is consistent with what you see in other enthusiast platforms where the automatic depreciates into ordinary-use buyers while the manual retains collector and specialist demand.

6. What this means for import buyers

You are buying into the enthusiast/export channel

If you are reading this as an import buyer, you are not the convenience-buyer segment. You are the specialist channel. In that channel:

• MT cars are preferred for off-road use, modification, and export market resale
• AT cars can be had at modest discounts at comparable mileage and condition in mid-range years, or at premiums for very recent low-mileage examples
• The price gap is not large enough to be the primary decision factor — year, mileage, condition grade, and rust provenance will move the price by more than transmission choice for most buying decisions

US buyers (2026 eligibility: 1型–3型 only)

Cars that became 25-year eligible by 2026 are 1998–2000 production year (1型, 2型, early 3型). In this range, both AT and MT examples exist in the Japanese market. Early-type AT examples in the ¥36万–¥55万 range (high mileage) and MT examples in the ¥38万–¥93万 range are both accessible.

For export to the US, MT is the stronger choice — the US market for imported kei SUVs favours manual transmissions, and MT early-type JB23s have clearer export-market demand from US buyers who want a tunable base. That said, if budget is the constraint, AT early types are cheaper on average at comparable mileage in the retail market.

For a full breakdown of what each US-eligible revision type offers mechanically, see JB23 revision types.

Canadian buyers (1型–8型 eligible)

Canadian buyers have access to a wider range including 5型–8型, where enthusiast demand for MT is strongest. In this range, the MT vs AT question matters more — a well-specified 7型 MT at ¥80万–¥100万 is a serious value proposition for an export buyer. AT 7型 examples in similar condition tend to list at similar levels in retail, but the buy-back evidence suggests MT holds value better over time.

Auction access matters more than transmission choice

Regardless of AT vs MT, your effective cost floor is set by the wholesale auction level — not retail asking prices. As detailed in researching Japanese auction prices, retail marks up by hundreds of thousands of yen above what exporters pay at auction. The transmission question is secondary to getting proper auction access. If you are paying retail for a JB23, the AT vs MT premium or discount is probably smaller than your retail-vs-auction gap.

For broader context on Japan’s export market dynamics, see the USS monthly auction index and Section 232 tariff considerations for Japan car imports.

7. Limitations

This section is mandatory. Use the data above with these constraints in mind.

N is small for a definitive AT vs MT comparison. CarSensor AT: 47 total listings, 19 sampled (1 duplicate removed). CarSensor MT: 32 total, 27 sampled (2 pre-JB23 entries excluded). Goo-net AT: 13 total, all captured. Goo-net MT: 16 total, all captured. This covers approximately 108 listings before deduplication. Both platforms likely list some of the same cars. We did not attempt to deduplicate across platforms, so unique-car N is lower.

Retail asking prices are not what you pay. Dealers set retail prices above auction-cleared prices. These figures are useful for relative comparison; they are not your landed cost. See researching Japanese auction prices for the methodology to estimate your actual cost floor.

No grade-stratified comparison. The single largest variable in JB23 pricing is not transmission — it is condition grade (clean title vs repair history) and mileage. Our sample mixes grades. A repair-history AT car and a clean MT car at similar mileage are not comparable, but they appear in the same dataset.

Auction data has limited MT coverage. The July 2026 auction data from usedcar-market-log.com shows AT prices for 2015–2017 but MT data only for 2017 and 2018. We cannot make a full year-by-year AT vs MT auction comparison from this source alone.

The buy-back claim about MT outperforming over time is qualitative. The hanamaru870 statement that “as cars age, MT valuations increasingly exceed AT” is a market observation from a buy-back service, not a price series. We include it as consistent with observed market behaviour, not as a measured finding.

Prices are volatile. JB23 prices have been subject to upward pressure from JB64 scarcity, US 25-year eligibility unlocks (2023–2025), and export demand. These factors can shift relative AT/MT pricing rapidly. Data collected 2026-08-20.

8. Method and sources

Data collection:

• CarSensor JB23 AT retail listings (47 total): https://www.carsensor.net/usedcar/freeword/%E3%82%B8%E3%83%A0%E3%83%8B%E3%83%BC+JB23+AT/index.html — retrieved 2026-08-20
• CarSensor JB23 MT retail listings (32 total): https://www.carsensor.net/usedcar/freeword/%E3%82%B8%E3%83%A0%E3%83%8B%E3%83%BC+JB23+MT/index.html — retrieved 2026-08-20
• Goo-net JB23 AT filtered listings (13 total): https://www.goo-net.com/cgi-bin/fsearch/goo_used_search.cgi?category=USDN&mission=AT&phrase=%E3%82%B8%E3%83%A0%E3%83%8B%E3%83%BC+jb23&query=%E3%82%B8%E3%83%A0%E3%83%8B%E3%83%BC+jb23 — retrieved 2026-08-20
• Goo-net JB23 MT filtered listings (16 total): https://www.goo-net.com/cgi-bin/fsearch/goo_used_search.cgi?category=USDN&mission=MT&phrase=jb23&query=jb23 — retrieved 2026-08-20
• USS auction settlement data, July 2026 (AT: 2015–2017; MT: 2017–2018): https://usedcar-market-log.com/market-price-jimny/ — retrieved 2026-08-20
• JB23 buy-back market analysis (qualitative AT/MT aging pattern): https://www.hanamaru870.net/column/jimny_satei/ — retrieved 2026-08-20
• JB23 type history and transmission availability by type: https://jimnylife.jp/jb23summary01-10/ — retrieved 2026-08-20
• Japanese-language market commentary (AT/MT ratio in used market): https://www.carseven.co.jp/satei/br160001/cn023 — retrieved 2026-08-20

What this analysis can establish: That JB23 retail prices for AT and MT overlap substantially at any given year and mileage band; that available auction data shows a modest AT premium in 2017 wholesale transactions (¥24万 gap at evaluation 5); and that market commentary from buy-back professionals consistently describes MT holding value better as cars age into enthusiast ownership.

What this analysis cannot establish: A statistically reliable AT vs MT price differential from retail data alone, given the sample size and uncontrolled variables (grade, mileage distribution, regional provenance, modification status). A buyer-specific cost estimate — your landed cost depends on auction access, exporter fees, freight, and destination costs that retail listing prices do not capture.