Itinerary
Planning Peru: The Country You Book Backwards
Most countries are planned forward: you decide roughly what you want to see, work out an order, and then book the pieces. Peru does not work like that, and trying to make it work like that is why so many Peru itineraries end up compromised. A small number of things here are capped, dated and sold out months in advance, and everything else has to be arranged around them. Getting that dependency right at the start is worth more than any amount of clever routing later.
Updated 2026-08-15
The chain, and why it runs backwards
The citadel at Machu Picchu is sold as a timed entry with a capped number of places, purchased in advance through the official system and not available on the day. That single constraint propagates outward. Your entry time determines which train you need, because the town below the site is reached only by rail and the journey takes hours. The train you need determines which night you must be in the Sacred Valley or in that town. And those nights determine when you fly into Cusco, which determines everything else.
So the correct order of operations is the reverse of instinct. Fix the citadel entry first. Then the train, in both directions. Then the accommodation those dates require. Then the flights. Then, and only then, decide what else the trip contains. Booking the train before the entry ticket is the specific mistake that produces the most misery, because trains and tickets sell out on different schedules and a train that does not match your entry slot is worth nothing.
If you intend to walk the Classic Inca Trail, the chain extends much further back. Trail permits are capped daily — a number that includes guides and porters, so the trekker allocation is considerably smaller than the headline figure — they are released annually, they must be bought through a licensed operator rather than directly, and the popular operators clear their allocation within hours of release. That means a trek is not something added to a Peru trip; it is the fixed point the entire trip is designed around, often the better part of a year ahead. The trail also closes for maintenance for a month each year, which is worth confirming against your dates before anything else.
The alternative treks — the several-day routes over higher passes, and the gentler community routes — are not permit-capped in the same way and can be arranged much closer to departure. If your dates are fixed and the Classic Trail is gone, they are genuinely good rather than consolation, and one of them reaches a higher and more dramatic pass than the Classic Trail does.
The altitude order almost everyone gets wrong
Here is a fact that reorders a Peruvian itinerary and that most visitors never learn: Machu Picchu is a thousand metres lower than Cusco. The citadel sits at around 2,430 metres and the town below it at around 2,040, while Cusco is at roughly 3,400. The Sacred Valley floor, in between, sits lower than the city as well.
The standard sequence — fly into Cusco, spend a few days acclimatising there, then go down to Machu Picchu — therefore starts at the highest point in the region on the day your body is least prepared for it. The better sequence runs the other way: fly in, transfer straight down to the Sacred Valley without spending the first night in the city at all, spend a day or two there among the valley sites at a gentler altitude, continue down the line to the citadel, and come back up to Cusco at the end when you are adapted. You see the same things in a more comfortable order, and you spend your first night several hundred metres lower than you otherwise would.
Two caveats keep this honest. The transfer from the airport down to the valley takes over an hour, which is a real ask immediately after a flight, though a considerably smaller ask than a bad night at 3,400 metres. And the high excursions sold from Cusco — the coloured-mineral ridge and the glacier lakes — go substantially higher than the city itself, so they belong at the end of a stay rather than near its start, whichever order you use for everything else.
How long the core takes
The Cusco region is the trip and everything else is an extension. Give it five days as an absolute minimum: the valley, the citadel with its travel days either side, and the city. Six or seven is comfortable and allows a day of slack, which this particular region genuinely needs. Add two days in Lima at one end, since you will pass through it anyway and the city is worth more than an airport transfer.
That makes a one-week trip Lima plus the Cusco circuit, and nothing else — which is a good trip and should not be apologised for. Ten days lets you add one extension. A fortnight lets you add two, or one extension done unhurriedly. Three weeks starts to make the north coast or a serious trek realistic.
The pacing rule that matters here is about altitude rather than distance: after a high day, take a low one. The itineraries that leave people exhausted are not the ones that cover the most ground but the ones that stack a mountain day on a mountain day without a valley in between.
The three extensions, and choosing between them
The southern circuit is the most popular addition: a white-stone colonial city with volcanoes behind it, a canyon where condors ride the morning thermals, and Lake Titicaca with its reed islands and weaving communities. It links by road and by air, it stays high throughout — the lake is higher than Cusco — and it suits travellers who want more of the Andes rather than a change of register.
The Amazon is the change of register, and it splits into two quite different propositions. The southern gateway is a short flight from Cusco and slots neatly onto the end of a Cusco trip, which is why most first visits use it. The northern gateway is a separate journey reached from Lima, deeper into the basin and served by the river rather than the road, and it suits a trip built around it rather than one adding it. Either way, treat lodge transfers as their own travel days and do not book a same-day onward flight.
The north coast is the one almost nobody does, and it is the strongest argument for a third week. It holds the great pre-Inca cultures — enormous adobe cities, decorated pyramids, and some of the richest tomb finds made anywhere in the Americas — with a fraction of the visitors of the Inca circuit and a warm dry climate year-round. If you have been to Peru before, or if crowds are what you most want to avoid, this is the answer.
Seasons, which do not agree with each other
The highlands have a dry season from roughly May to September and a wet one from about November to March. Dry means clear skies, reliable trekking and the best chance of a cloudless citadel — and it is also the peak, when tickets, trains and permits are hardest to get and prices are at their highest. The wet season is greener, far quieter and substantially cheaper, with the trade-off that mountain views come and go and trails are muddy. The shoulder months at either end are the reliable compromise.
The coast runs on the opposite calendar and this surprises people. Lima's summer, from about December to March, is its sunny warm season; its winter is cool, damp and sits under a persistent grey sea mist for months. So the months that are best in the mountains are the greyest in Lima, and the months that are pleasant on the coast are the wettest in the Andes. A trip covering both is always a compromise, and the shoulder months serve both better than either extreme.
The Amazon is wet year-round and the meaningful distinction is water level rather than rainfall. High water floods the forest and lets boats travel among the trees, reaching areas otherwise inaccessible; low water exposes beaches and river banks and makes walking trails usable. Neither is better — they are different experiences, and a lodge will tell you which applies to your dates.
What goes wrong
The commonest failures are all versions of the same thing: treating a capped, dated booking as though it were flexible. Booking the train before the citadel ticket. Assuming a same-day ticket might be available. Leaving the trek permit until the trip is otherwise planned. Building the Amazon leg so that its transfer day collides with the flight home.
The second cluster is altitude, and it is covered above: arriving into Cusco and going straight out sightseeing, or sequencing the region so the highest ground comes first. What altitude actually does to you, and what genuinely helps, is set out in the health and safety guide. The third is under-budgeting the Cusco region as a whole — three days there is not enough for what people expect to get out of it, and the days lost to the train journeys are real days.
Finally, protect the end of the trip. Peru's roads and rail can close at short notice for demonstrations or weather, domestic flights out of mountain airports are weather-dependent in the afternoons, and the citadel corridor is exactly where a disruption becomes a missed international connection. Keep a buffer day in Lima before flying home rather than routing straight from Cusco to an overnight international departure, check the current travel advisory before and during the trip, and treat that spare day as the cheapest insurance in the itinerary.