For a five-day Shikoku trip, I would choose Takamatsu when two or more full days point toward Naoshima, Shodoshima, Kotohira or eastern Kagawa; I would choose Matsuyama when Dogo Onsen, Matsuyama Castle and western Ehime are the trip’s main event. If both sides contain genuine must-dos, the best answer is usually not “pick the more convenient city.” It is one deliberate hotel move.
The reason is measurable. Takamatsu and Matsuyama are connected by JR Shikoku’s Limited Express Ishizuchi, but the intercity segment is roughly a 2.5-hour planning block rather than a suburban hop. A same-day out-and-back can consume around five hours of train time before you add hotel-to-station movement and local transport at the destination. Avoiding one hotel change is not a win if it creates that journey twice.
Fact-checked August 26, 2026. JPY remains the primary currency; approximate USD figures use about ¥159 = US$1 as a working reference. Timetables, ferry operations, fares and pass rules can change.
If Takamatsu wins, the next local decision is Takamatsu Station or Kawaramachi. If you are trying to add Naoshima to several Kagawa days, compare the longer Takamatsu itinerary before inserting Matsuyama simply because it is another major Shikoku city.
Takamatsu or Matsuyama: the 30-second route decision
↔ Scroll table left or right
Your five-day trip | Base I would use | Why |
Naoshima + Shodoshima + Kotohira/Ritsurin | Takamatsu for all four nights | Port mornings and eastern-Kagawa rail trips recur. Matsuyama would add a long transfer without fixing a daily problem. |
Dogo Onsen + Matsuyama Castle + another western-Ehime day | Matsuyama for all four nights | The best Dogo hours are in the evening, and repeated travel back east would destroy that advantage. |
Naoshima or Shodoshima plus Dogo and Matsuyama Castle are all non-negotiable | Split 2+2 or 3+1 | One hotel move replaces a repeated long intercity return. |
Takamatsu arrival, Matsuyama departure — or the reverse | Split strongly favoured | An open-jaw route removes the final backtrack. |
Same airport in and out, with an early flight on Day 5 | Protect the final-night side | The cheapest airfare can become expensive in time if the last hotel is across Shikoku. |
You also insist on Iya or Kochi | Cut a region before adding a third base | Five days is too short for three one-night hotel clusters without turning the trip into a transport sample. |
ValueNavi true time-cost test: one hotel move versus one same-day return
The cleanest way to decide is to compare the cost of the hotel move with the cost of refusing it. A suitcase move feels annoying because you notice it. A long train each way feels less dramatic because it is spread across the day, but it can cost much more usable holiday time.
↔ Scroll table left or right
Choice | Current transport cost | Time cost | What it means |
One Takamatsu → Matsuyama JR move | Regular non-reserved fare about ¥6,160 ≈ US$38.70; normal-period reserved fare ¥6,690 ≈ US$42.10 | Roughly 2.5 hours, then local transport | Reasonable as a single linear transfer because the journey replaces a hotel base. |
Takamatsu → Matsuyama → Takamatsu in one day by regular JR fares | About ¥12,320 ≈ US$77.50 using non-reserved fares | About five hours of intercity rail before local movement | This is the hidden cost of “one hotel at all costs.” |
One-way JR with the current e5489 Toku-wari WEB ticket | ¥4,200 ≈ US$26.40 for a reserved seat | Train time does not shrink | Excellent for a planned base move; it makes a same-day return cheaper, not shorter. |
Highway bus one way | ¥4,400 ≈ US$27.70; current round-trip ticket ¥8,400 ≈ US$52.80 | Roughly 2h25–2h50 depending on service, before traffic risk | Saves money versus regular JR, but does not turn the two cities into one practical base. |
JR Shikoku’s 2026 Toku-wari WEB ticket is particularly useful for this itinerary. The Takamatsu–Matsuyama reserved-seat fare is ¥4,200 instead of the normal ¥6,690 during the current offer period, and it must be bought by the day before travel. If your date and booking conditions fit, that lowers the monetary penalty for splitting the trip. It does not change the main conclusion: the reason to split is time, not the headline rail fare.
For two adults, a regular non-reserved same-day rail return would be roughly ¥24,640 (about US$155) in total. At that point, keeping one hotel because “moving bags costs money” deserves a proper comparison with the actual luggage-transfer, taxi or locker cost. The transport bill can be much larger than the cost of making one base change intelligently.
Takamatsu earns its nights by protecting ferry mornings
Takamatsu’s biggest advantage is not that it is a larger or more interesting city. It is that the port is part of the morning routine. Shikoku Kisen lists JR Takamatsu Station at about a five-minute walk from Takamatsu Port. That turns an island day into “hotel → port → boat,” rather than “hotel in another city → two-and-a-half-hour intercity journey → port → boat.”
For Naoshima, the current Takamatsu–Miyanoura timetable includes a 07:20 high-speed passenger boat arriving 07:50 and an 08:12 ferry arriving 09:02. The adult one-way fares are ¥1,590 ≈ US$10.00 for the high-speed boat and ¥680 ≈ US$4.30 for the ferry. The operator says boarding reservations are not accepted; passengers are handled first-come, and ticket sales at Takamatsu begin 30 minutes before high-speed departures and 40 minutes before ferry departures.
That is exactly the sort of schedule where a hotel near Takamatsu Station/port buys real itinerary value. A 07:20 boat is not just “thirty minutes earlier than the ferry.” It can move your whole island day forward before later arrivals build. If Naoshima is one of the reasons you came to Shikoku, I would rather protect that morning than protect the convenience of never changing hotels.
Shodoshima makes the Takamatsu case even stronger — but 2026 service changes matter
The current Takamatsu–Tonosho ferry timetable starts even earlier, with ferries at 06:25 and 07:20, taking about 60 minutes. Passenger fare is ¥700 one way ≈ US$4.40. The high-speed boat currently starts at 07:40, takes about 35 minutes and costs ¥1,400 one way ≈ US$8.80.
There is an important 2026 failure mode. Shodoshima Ferry reduced the Takamatsu–Tonosho high-speed service from 15 departures a day to eight from May 16, 2026, citing fuel-price pressure, and says the reduction is in effect for the time being. The current fast-boat timetable therefore matters more than an older blog post that says there are frequent departures all day.
For foot passengers on the high-speed boat, reservations are not accepted and boarding is first-come. Sea conditions can also disrupt services. I would not design a five-day itinerary that depends on a razor-thin ferry-to-airport or ferry-to-long-distance-train connection on the final day. Takamatsu is a strong island base partly because it gives you more recovery options when the marine timetable changes.
Takamatsu Station or Kawaramachi? The answer depends on which friction repeats
Once Takamatsu wins the city-level decision, you still have a local-base decision. If Naoshima or Shodoshima appears on two separate mornings, I would stay around Takamatsu Station/port. That keeps luggage, JR departures and ferries in the same small operating zone.
If the trip is mostly Ritsurin Garden, udon, shopping arcades and city evenings with only one easy island day, Kawaramachi can be the better lifestyle base. The important point is that “Takamatsu” is not one homogeneous hotel location. Use the Takamatsu Station vs Kawaramachi guide after you know how many ferry mornings you are actually protecting.
Matsuyama earns its nights by protecting the Dogo evening
Matsuyama works differently. Its biggest accommodation value is often not an early departure but the ability to stop travelling. Matsuyama Castle is a daytime sight; Dogo Onsen, dinner, the shopping street and a ryokan stay can continue into the evening. If the emotional centre of the western half of the trip is “finish the day in Dogo and stay there,” forcing a return to Takamatsu makes the plan worse even if the train connection is technically available.
Do not count arrival at JR Matsuyama Station as arrival at Dogo. The onsen district is a separate local-transport leg, and local hotel choice changes the trip. A station-area hotel protects an early intercity departure. A Dogo stay protects the onsen evening. If you are only taking a quick bath and leaving, Dogo may not justify a hotel change; if the evening itself is part of the reason for visiting, it often does.
The “direct train” test: direct does not mean commute-worthy
JR Shikoku’s Limited Express Ishizuchi connects Takamatsu and Matsuyama. That makes a one-time base move easy to understand. It does not make a daily return efficient. The distinction matters because route planners often label a journey “direct,” which psychologically makes it feel simple even when the duration is long.
For a five-day trip, I use this rule: if the distant side has only one attraction, question whether it belongs in the itinerary at all. If it has an evening plus a second meaningful sight or day, move the hotel. The worst middle ground is keeping the distant side because it matters, then refusing the hotel move because the train is direct.
2+2 or 3+1 nights? Give the extra night to the side that protects two full days
A five-day/four-night split does not need to be symmetrical. Takamatsu 2 + Matsuyama 2 is clean when each side has one strong full day plus meaningful arrival/departure time. Takamatsu 3 + Matsuyama 1 makes more sense when Naoshima and Shodoshima are separate full days and Matsuyama is primarily one castle-and-Dogo finish. Reverse that logic if western Ehime carries more of the trip.
↔ Scroll table left or right
Split | Best fit | Hidden risk |
4 nights Takamatsu | Two islands + Kagawa | Adding Matsuyama as a day trip can turn a good eastern route into a five-hour transport day. |
3 Takamatsu + 1 Matsuyama | Island-heavy trip with one Dogo/Matsuyama finish | The single Matsuyama night only works if arrival is early enough to use the afternoon/evening. |
2 + 2 | At least one full-day priority on each side | Do not waste the transfer day on a third remote excursion. |
1 Takamatsu + 3 Matsuyama | One Kagawa/Setouchi taste, then western Ehime | A single Takamatsu night is poor if you are trying to squeeze both Naoshima and Shodoshima into it. |
4 nights Matsuyama | Dogo, castle and western-Ehime focus | Takamatsu should become a gateway, not a compulsory sightseeing checkbox. |
The extra night should protect the side with two full mornings or two full days, not whichever city seems more famous. Hotel nights are operating assets: they should improve what happens before 10:00 or after 18:00, when a day-tripper has the least flexibility.
Use the base-move day for city content, not another remote excursion
The most resilient Takamatsu → Matsuyama transfer day is deliberately boring in transport terms. Have breakfast, check out, take the intercity train or bus, leave luggage, then use the afternoon for Matsuyama Castle and the evening for Dogo. Alternatively, keep a gentle morning in Takamatsu, move after lunch and make Dogo the only commitment.
What I would not do is check out, take an island ferry, return to Takamatsu, collect bags, cross Shikoku and then expect a relaxed Dogo dinner. That plan has too many independent clocks: ferry, luggage collection, intercity departure and hotel check-in. One delay pushes every later step.
Open-jaw flights can make the split almost self-solving
If airfare works, Takamatsu-in and Matsuyama-out is the cleanest direction for an east-to-west route; Matsuyama-in and Takamatsu-out works just as well in reverse. The important advantage is not the airports themselves. It is deleting the final cross-island backtrack.
If both flights use Takamatsu Airport, however, a final Matsuyama night can become a bad choice when Day 5 has an early departure. The same applies in reverse. Price the flight and hotel route together. A flight that saves ¥3,000 but forces an extra intercity night move or a stressful early-morning crossing is not necessarily cheaper in travel value.
2026 rail-pass trap: the five-day pass disappears on October 1
This is where older Shikoku planning advice is especially easy to get wrong. As of August 26, 2026, the ALL SHIKOKU Rail Pass still has a five-day version for purchases through September 30: ¥17,000 ≈ US$107 for overseas sales and ¥17,500 ≈ US$110 for domestic/internet sales. It covers non-reserved JR Shikoku limited expresses and, under the current rules, the Takamatsu–Tonosho Shodoshima Ferry plus Shodoshima Olive Bus. Iyotetsu in Matsuyama has already been excluded since March 7, 2026.
From October 1, 2026, that changes materially. JR Shikoku has announced that new ALL SHIKOKU Rail Pass sales will offer only three-day and seven-day versions; the four-day and five-day passes will be discontinued. For passes bought from October 1 onward, Shodoshima Ferry and Shodoshima Olive Bus will also no longer be included. Passes purchased by September 30 retain those Shodoshima benefits under the official transition rule.
↔ Scroll table left or right
Travel-planning date | Relevant rule | Decision impact |
Pass purchased by Sep 30, 2026 | 5-day pass still exists; eligible Takamatsu–Tonosho ferry and Olive Bus coverage remains | Could fit an island-heavy five-day route, but calculate your actual covered rides. |
Pass purchased from Oct 1, 2026 | Only 3-day and 7-day versions; Shodoshima ferry/bus coverage removed | Do not buy a seven-day pass simply because your trip lasts five days. |
Matsuyama local transport | Iyotetsu not included since Mar 7, 2026 | Budget local tram/bus costs separately; old pass guides may be wrong. |
There is another useful break-even check. A regular Takamatsu–Matsuyama non-reserved return is about ¥12,320, still below the ¥17,000 five-day pass by itself. The pass only becomes compelling when enough additional covered JR journeys — and, for eligible pre-October purchases, covered Shodoshima transport — close that gap. Trip length is not the break-even calculation; covered transport is.
Three five-day routes I would actually book
Route A: Setouchi and Kagawa — four nights in Takamatsu
Best for: Naoshima + Shodoshima + Kotohira/Ritsurin. Keep the hotel stable. Put the two ferry-dependent days where weather and museum opening days make sense, then use Kagawa as the flexible land-based day. Matsuyama is omitted, not “saved for a quick day trip.” That omission is what keeps the route strong.
Route B: both sides genuinely matter — two nights Takamatsu + two nights Matsuyama
Best for: one strong Setouchi/Kagawa day plus a real Matsuyama/Dogo stay. Use one side for Days 1–3, make the intercity move once, then finish on the other side. If flights are open-jaw, align the direction with them. The move day should contain city sights, not another long excursion.
Route C: western Shikoku — four nights Matsuyama
Best for: Dogo, Matsuyama Castle and another Ehime priority. Takamatsu can be removed entirely or treated only as an arrival gateway if the flight or rail fare requires it. The itinerary becomes stronger when it stops pretending that every major Shikoku city must appear.
Families and older travellers: choose between luggage friction and repeated journey friction
There is no universal rule that families should avoid hotel changes. With young children or older relatives, packing once is attractive, but five hours of repeated intercity travel can be more tiring than one controlled move. The right question is which friction is larger for your group.
If Matsuyama is only “maybe a castle photo and one bath,” I would delete it and keep Takamatsu. If Dogo is a major family priority, move once and let everyone stop travelling after dinner. If mobility is limited, reduce the number of bases but also reduce the geographic ambition. Keeping every destination while eliminating hotel moves is usually the worst combination.
Weather and disruption: the itinerary needs one land-based fallback
Island-heavy Takamatsu itineraries have a risk that Matsuyama-heavy ones do not share to the same degree: marine services are timetable- and weather-sensitive. Shodoshima’s 2026 fast-boat reduction is a reminder that frequencies can also change for operational reasons, not only storms.
For a five-day trip with two island days, keep at least one land-based Kagawa day movable. Ritsurin Garden, Kotohira or central Takamatsu can absorb a ferry-day reshuffle better than a tightly pre-booked cross-island transfer. Do not put your only viable Naoshima ferry day immediately before a fixed early flight unless you accept the risk.
Red-team: when splitting Takamatsu and Matsuyama is the wrong answer
↔ Scroll table left or right
Condition | Why the recommendation flips |
Matsuyama has only one “nice to have” sight | Delete Matsuyama rather than paying a hotel move for a weak priority. |
Naoshima and Shodoshima are both must-dos | Protect Takamatsu nights; one Matsuyama night may steal the weather flexibility your ferry days need. |
You already added Iya or Kochi | A third region can turn a clean two-base plan into three fragmented one-night stays. Cut before adding. |
Very early round-trip flight from one side | The final night should move back toward that departure unless you change the flight. |
Your group finds repacking much harder than sitting on trains and the distant side is essential | A one-base plan can still be rational — but treat the repeated five-hour transport block as a conscious trade-off, not free convenience. |
Final verdict: move hotels only when the move deletes a larger repeated journey
Takamatsu is the better base for a five-day Shikoku trip built around Setouchi islands and Kagawa. Matsuyama is the better base for a trip built around Dogo and western Ehime. If both sides contain at least two meaningful blocks of travel value, split once.
The decision is not about which city has the better station, the cheaper hotel or the longer attraction list. It is about protecting scarce mornings and evenings. Takamatsu protects ferry mornings. Matsuyama protects the Dogo evening. A single intercity move costs roughly 2.5 hours; refusing that move can turn the same corridor into a five-hour return.
That is the rule I would use for the whole itinerary: change hotels only when one controlled move removes a larger repeated transport cost. If it does not, stay put. If it does, stop treating “one hotel” as automatically more convenient.
Your next decision
If Takamatsu wins, compare Takamatsu Station vs Kawaramachi for ferries, trains and evenings. If you are still trying to fit too many regions into four full days, use the four-day Shikoku route as a constraint test: the same principle applies even outside autumn — cut geography before adding hotel churn.
Okayama Korakuen and Kurashiki Day Trip: Foliage, Lights and Hotels
Where to Stay for Tokushima Awa Odori: Tokushima, Naruto or Takamatsu
Best Airport for Japan Autumn Leaves: Choose by Foliage Window and Open-Jaw Route
References
JR Shikoku: Limited Express Ishizuchi
JR Shikoku: Limited Express fare information
JR Shikoku: 2026 Takamatsu–Matsuyama Toku-wari WEB ticket and regular-fare comparison
JR Shikoku Bus: Bocchan Express Takamatsu–Matsuyama timetable and fares
Shikoku Kisen: Takamatsu–Miyanoura (Naoshima) timetable, fares and boarding information
Shodoshima Ferry: Takamatsu–Tonosho ferry timetable and fares
Shodoshima Ferry: Takamatsu–Tonosho high-speed boat timetable and fares
Shodoshima Ferry: May 2026 high-speed service reduction notice
JR Shikoku: ALL SHIKOKU Rail Pass terms effective March 7, 2026
JR Shikoku: ALL SHIKOKU Rail Pass changes from October 1, 2026
Transport schedules, discount-ticket sales periods, ferry operations, pass coverage and fares can change. Recheck the official operator for your exact travel date before paying for flights, passes or non-refundable hotels.