Strategy № 14

When to Upgrade Your Second Settlement to a City

Upgrading a settlement is a one-VP move that costs five resources. The timing matters more than the upgrade itself.

By Henrik Mølmer · Editor & Founder, Cartographer's Almanac

9 min read · May 17, 2026

TL;DR

Your first city is almost always correct as soon as you can afford it. The second city is a timing decision: build too early and your settlement-expansion stalls, build too late and you give up two-resource production for too many turns. The right window is usually when you've already secured one more settlement spot and have surplus wheat + ore arriving in the next 3-4 turns.

The cost-benefit, plainly

Upgrading a settlement to a city costs 2 wheat + 3 ore. The benefit: that intersection now produces 2 resources per activation instead of 1. The VP swing is +1.

That's a steep cost for one VP. The actual reason to build cities is the production multiplier, not the VP. A city on a 5-pip hex produces 10 expected resources per 36 rolls instead of 5 — that's a structural production boost that compounds for the rest of the game.

So when to time the second one?

The window: usually VP 5-6

The standard build order: two settlements (VP 2) → first city upgrade (VP 3) → third settlement (VP 4) → second city upgrade (VP 5) → fourth settlement (VP 6) → from here, depend on dev cards and bonus VPs.

The second city slot in that order is around VP 5. It can shift earlier (VP 4) if you have a high-pip settlement and your dev card strategy hasn't started yet; it can shift later (VP 6) if you're focused on aggressive expansion or have a Longest Road race in motion.

When to build the second city early

1. Your two settlements are both on high-pip hexes

If both settlements are 14+ pips, the city multiplier is most valuable. A city on 16 pips produces 32 per 36 rolls. The compounding edge over players with city-on-8 production is substantial.

2. Expansion is blocked

If your road network is constrained (coast, opponent, terrain), you can't build a third settlement at strong pips. The next-best move is the second city upgrade — it converts surplus into production without needing a settlement slot.

3. Wheat and ore are accumulating

If you produce wheat and ore on red numbers (6, 8) and they're piling up in your hand (risking 7-roll discards), upgrading converts the surplus into board state. Convert hand to city before the discard event hits.

When to delay the second city

1. The third settlement is more valuable

If you can settle at a 10+ pip intersection on a new resource, the third settlement adds more total expected production than the second city upgrade. Settlements: 1 VP + 1 production per hex. City upgrade: 1 VP + 1 production per existing hex (doubles existing). The new hexes from a settlement usually win when the alternative city's pip total is in the single digits.

2. Longest Road race

You can't both upgrade and chase Longest Road with the same resource pool. The road race needs wood and brick — distinct from city resources. But the tempo of "upgrade vs. build" matters: upgrading uses a turn you'd otherwise have spent extending.

3. Pre-empting the close

If you're at VP 5 and a dev card hoard could put you at 7+ within two turns, save the wheat and ore for dev cards instead. Dev cards have a chance of being hidden VP — better expected VP per resource than a city upgrade at certain card counts.

The specific decision points

At VP 4 (one city, one settlement, plus original two)

Default: build the third settlement. Exception: if the second settlement is on a 15+ pip corner and the third settlement option is sub-10 pips, upgrade instead.

At VP 5 (one city, two settlements)

This is the typical second-city window. If both your remaining settlements are 10+ pips, upgrade the higher one. If one is 8 pips and one is 14, upgrade the 14. If both are sub-8 pips, consider whether a fourth settlement is feasible first.

At VP 6 (one city, two-three settlements)

You're entering endgame mode. Cities at this point are slow — they take a turn to build and the game may end in two turns. Switch to dev cards unless the city upgrade is part of a known closing sequence.

The double-city compounding effect

Two cities at 14 pips each produce 28 expected resources per 36 rolls between them. One city + one settlement at 14 pips each produce 14 + 28 = 42 — but with the upgrade cost paid. The break-even on the upgrade investment (2 wheat + 3 ore = 5 resources) is roughly 5-7 turns of accelerated production. That's fast in Catan terms, which is why cities are the most efficient resource conversion in the game.

The decision in practice

Before each turn, ask: "If I had 2 wheat and 3 ore right now, what would I do?" If the answer is "upgrade my best settlement," the second city is in your near-term build plan. If the answer is "save for a dev card" or "trade for wood + brick to settle a fourth," then upgrade isn't the next move.

Test the math on a generated board: build a Catan layout, walk through a few opening turns, and compute when your wheat + ore arrive. The second-city window usually opens around turn 8-10 on a balanced opening.

❦

Filed under

strategy cities timing mid-game