Years ago, before Makasa existed, Ngomi and I drove into a herd of wildebeest with no end to it. The animals bolted off in short stiff jumps as we came through, and then the Land Cruiser got a flat tyre. The guide switched the engine off and the silence of the Serengeti came down on us. Wildebeest to the left, to the right, in front, behind. He got out to change the wheel, and the animals let themselves be bothered less and less, and the herd closed in around the car. I can still hear the snorting.
The migration is not a performance with a start time. It is a loop of around 800 kilometres, about 500 miles, that never stops moving. It follows the rain rather than the calendar, and it takes no notice at all of when you have booked your leave.
So this is where the herds are and when, what the science does and does not know, and why calving season in February leaves more of an impression on a lot of our guests than the crossings in August do. And at the end, what the crowds have done to this trip, because an honest answer has to include that.
When is the great migration in Tanzania
The migration runs all year. The herds are always moving somewhere inside the Serengeti and Mara ecosystem. So the question is not when the migration happens. It is where the animals are standing during the weeks you can travel.
Roughly, it looks like this.
- December to March. The southern short-grass plains around Ndutu and Kusini, partly inside the Ngorongoro Conservation Area. Calving season.
- April and May. The herds shift northwest along the Moru Kopjes and west of Seronera. The long rains.
- May to July. The Western Corridor, towards the Grumeti river. The first crossings.
- July to October. The northern Serengeti, Kogatende, the Lamai Wedge, and the Masai Mara over the border. The Mara river crossings.
- October and November. Back south along the eastern side, past Bologonja and Lobo, as the short rains start.
That is a guideline and not a timetable. With early rain the herds stand further south in October than you would expect. In a dry year they hang around the Mara longer. In the weeks before you fly we ring our guides and the camps in the north to hear where the animals are that week, and we move the route to match.
What sets the herds moving
They move for grass, and for what is in the grass.
The southern plains sit on volcanic ash blown off the Ngorongoro highlands. The short grass growing on it is high in calcium, phosphorus and nitrogen, which is what a pregnant or nursing female needs. Research by Holdo and Fryxell describes two gradients running against each other: rainfall increases towards the northwest, soil fertility increases towards the southeast. The herds shuttle between the better grass in the south and the water in the north.
Once the southern grass is grazed off and the waterholes dry up, there is no choice left in it. They go.
How they find rain hundreds of kilometres away is less well understood than people assume. Wildebeest have a good nose and they can see a storm building from tens of kilometres off. Elephants have been shown to react to distant rain and lightning. For wildebeest it is a likely explanation rather than a proven one.
Which is why every month-by-month chart, including the one further down this page, is roughly right and never exact. The migration follows the rain.
How many animals are in it
The most recent official count comes from TAWIRI, the Tanzania Wildlife Research Institute, working with TANAPA, the Frankfurt Zoological Society and the University of Glasgow. The fieldwork was flown in March 2023 and published in 2024. It put the wildebeest at 1,366,109, with a margin of error of more than 230,000 animals. Alongside them travel around 200,000 zebra and some 400,000 Thomson’s gazelle, with Grant’s gazelle and eland moving with the herds in numbers nobody has counted.
The zebra usually walk at the front. They take the taller, coarser grass, which opens up the shorter grass underneath for the wildebeest coming behind them.
In September 2025 a team at the University of Oxford led by Isla Duporge used satellite imagery and image recognition, the first count built on satellite data, and came out at fewer than 600,000 wildebeest. The Glasgow researchers Grant Hopcraft and Colin Torney were not convinced. Only part of the range was photographed, they argued, and up to a fifth of the animals can be standing under trees where a satellite will not see them. The Oxford authors say themselves that they are not claiming an exact figure.
We work with a range of 1.3 to 1.5 million, and we tell you that scientists are arguing about it. A round number that sounds good is worth less than one you can lean on.
The Serengeti migration, month by month
December, January, February. The herds come down onto the southern plains around Ndutu, on the edge of the Ngorongoro Conservation Area. Short grass, open views, and a great many predators. This is the most underrated season of the year, and calving gets its own section further down.
March and April. The long rains usually start from about the middle of March. The mega-herd breaks into smaller groups and those groups move north through acacia woodland. The rain comes in showers, late afternoon or at night, not all day, and the morning drives are usually clear. The roads get heavy, the camps get emptier, the country is green and the skies are enormous. No guest has ever come back and told me they regretted March.
May and June. The move west begins, past the Moru Kopjes, then the country west of Seronera, then into the Western Corridor. They cross the Mbalageti river on the way. By June the animals are gathering on the south bank of the Grumeti, waiting until the group is big enough.
June and July are the rutting months. Bulls hold temporary territories in the middle of a moving herd and defend them against rivals while the whole mass keeps walking. The males are worn out and barely eat for weeks. The noise is deafening, and almost no travel description mentions it.
The Grumeti crossing is smaller than the Mara one. The river is narrower, with deep permanent pools holding some of the largest Nile crocodiles in Africa, and there is far less traffic than up north. That is why I recommend this stretch more often than you might expect.
July, August, September. The herds reach the north, around Kogatende and the Lamai Wedge, and start crossing the Mara river. Statistically, August and the first days of September are the best weeks, and they are also the busiest. The animals cross several times in a season, back and forth, depending on where a shower has put fresh grass.
October. The animals move back and forth between the northern Serengeti and the Masai Mara, so crossings happen in both directions. A southbound crossing draws fewer vehicles and is often just as good to watch.
November and December. The short rains start and the herds move south again along the eastern side of the Serengeti, back towards Ndutu. The circle closes.
Calving season in Ndutu, January and February
This is the part of the migration most people skip, and it is the part we would rather be standing in ourselves.
Around 500,000 calves are born, almost all of them inside a window of two to three weeks, with the peak usually falling in February. About 80 per cent of the females give birth in that same window, and on the busiest days there are thousands of births in a day, spread out across the whole plain.
Why nearly all the calves are born at once
That timing is not coincidence. It swamps the predators with more prey than they can physically eat. A calf born two weeks early, or two weeks late, is walking on its own among hungry hyenas and lions. A calf born at the peak disappears into the herd.
A newborn wildebeest is on its feet within a few minutes and keeping up with the herd inside a day. On open ground there is nowhere to hide, so running is the only defence it has. You watch mothers shoving their calves upright while they can barely stand themselves.
And the calcium-rich soil at Ndutu is why it happens there and not a hundred kilometres away, around sixty miles, where the ground is poorer. The mineral goes from the soil into the grass and from the grass into the milk, and the calf builds its skeleton out of it in a few weeks.
Why February is the strongest month for cheetah
Cheetah hunt by day, which lions and leopards mostly do not. They find prey by eye and then sprint, and the sprint is rarely longer than 300 or 400 metres, a quarter of a mile at most, because the energy cost of going further is too high. On the open plains at Ndutu, with thousands of inexperienced calves standing about, the sprint rarely needs to be long.
Which means you can watch the whole sequence in daylight. The choosing, the walk in, the sprint, the catch. Usually without the vehicle having to move at all.
After a kill a cheetah has to lie down before it can eat, because its body temperature climbs so far during the sprint. That is the moment it is most vulnerable, and you are watching it happen.
Why you get closer at Ndutu than in the Serengeti
Ndutu sits formally inside the Ngorongoro Conservation Area and not inside Serengeti National Park. In the conservation area you are allowed to drive off the tracks. In the national park you are not.
So a guide can follow a hunting cheetah across the plain and put the vehicle where the light is. That is why February produces the photographs people are still coming back to years later.
What it does to people
Calving season is hard to watch. Not every calf makes it, and I say that to guests beforehand so nobody is caught out by it.
What I notice is that people come back into camp quiet at the end of a day like that. Not sad. Full.
One practical thing. Stay in the Ndutu area rather than the central Serengeti, or you spend hours driving every morning before you get to anything. In January and February the plain is still green, the mornings are warm, it is far quieter than August, and it costs less.
The Mara river crossing, July to October
The herds collect on the bank, sometimes for hours, sometimes for days. Animals walk down to the water, turn round, come back up. The zebra hesitate longest, and the wildebeest will not go while the zebra are holding back, so the zebra are the trigger. Then one animal jumps, and the rest follow it in.
Sometimes they do not cross at all and drift back inland, and that is normal rather than a failure. We once stood at that bank for six hours and nothing happened. It was still one of the strongest days of that trip.
The crossing is not the border with Kenya
This is the most stubborn misconception about the migration, and most websites keep it alive. The herds go into Kenya at the crossing, the story goes, and come back later.
That is not how it works.
The Mara river runs straight through the northern Serengeti, well inside Tanzania. At Kogatende the animals move from the south bank to the north bank, and both banks are Tanzanian. The country north of the river is the Lamai Wedge, a triangle of rolling grassland between the river and the Kenyan border, and it is still inside Serengeti National Park. So the crossings at Kogatende happen entirely inside Tanzania.
The border with the Masai Mara lies a good way further north again. Part of the herds does cross into Kenya later in the season. Another part stays on the Tanzanian side the whole time and crosses the Mara back and forth several times, because each time the other bank has the fresh grass on it. You can watch crossings for weeks without ever going near Kenya.
Which means two things for you. You need no second visa to see a crossing, and you are not tied to the window when the animals happen to be in Kenya.
Which bank you watch from
The crossing points along the Mara are numbered. For each one it is fixed how many vehicles may stand there, how far back from the bank they have to stay, and which side of the river they stand on. At nearly every point only one bank is in use and the other stays empty.
Those rules came in as the area got busier, to protect the place and the animals. And they make a practical difference, because a herd that sees vehicles on both banks turns back more often than it crosses.
The guides who drive there every day know which point is open that morning and which bank applies. You do not need to get into any of that. Your guide knowing it is the part that matters.
What you do not see happening
Research by Subalusky and colleagues, published in PNAS, put the number of wildebeest drowning in the Mara at an average of around 6,250 a year, roughly 1,100 tonnes of animal. That carcass load feeds the river. Fish eat off it, vultures and marabou storks clean the banks, and the bones go on releasing phosphorus into the water for seven years. The river runs on the migration the way the migration runs on the grass.
If you want the practical side, where the crossing points near Kogatende are, how to improve your chances and how many days you realistically need, that is in our separate piece on the Mara river crossing.
Tanzania or Kenya, where do you see the migration
The migration is visible in Tanzania all year round. Kenya has a window of a few months, and even inside that window a large part of the herds is still on the Tanzanian side.
The word migration is itself a bit misleading. More than a million animals never stand together in one place. They move in dozens of separate groups across hundreds of square kilometres, each group going its own way. In August one herd is grazing in the Masai Mara, a second is still in the Lamai Wedge, and a third is crossing the Mara back to the south.
The country north of the Mara river, which is still Serengeti, usually holds the largest share of the migration from July to November, and in the dry months it often holds more animals than the Kenyan side does. We see it every year, and the camps up there report the same.
Two things keep the opposite belief alive. The Masai Mara is small and well connected and has been the backdrop for wildlife documentaries for decades. And Kenya has marketed the migration as a Kenyan product for a very long time. Good marketing, weaker geography.
None of which makes Kenya a lesser choice. It is a fine destination and the Mara is a splendid reserve. The point is only that if you are travelling in Tanzania you are not missing the migration, because the animals are here all year, including in the months Kenya markets hardest.
Since July 2024 the pricing has also pulled the two apart. Entry to the Masai Mara National Reserve runs at 200 US dollars per adult in high season, valid for twelve hours. The Serengeti is 70 dollars per person per 24 hours in high season and 60 in low season, with 18 per cent VAT on top. That gap is one of the reasons the northern Serengeti has got busier.
Which brings me to the part of this subject that safari websites tend to leave out.
What we saw on the ground last year, against what the calendar promised
The route held up. The herds were where we expected them. Calving fell in February and the river crossings happened in August.
What had changed was the crowding.
TANAPA figures reported in the Tanzanian press in May 2026 put international visitors to the Serengeti at about 388,000 in 2023, over 430,000 in 2024 and over 491,000 in 2025. Bed capacity inside the park now sits above 5,100, which the IUCN describes as a sixfold increase in ten years. In peak season there are hundreds of vehicles working the northern section on any given day.
In July 2025, at Kogatende, the official maximum at one crossing point is 60 vehicles and one of our guides counted more than 150. People had got out and were sitting on their bonnets. The herds were blocked and did not cross. TANAPA suspended six guides for a year, from every national park in the country and not only the Serengeti, and the Conservation Commissioner, Musa Kuji, called it a breach of professional ethics.
And at the Nyumbu crossing at Kogatende, on 30 August, a safari vehicle went into the Mara while its driver was racing the others to the bank. Four Swiss guests and the driver got out of a river full of crocodiles unhurt. TANAPA issued a warning afterwards about speeding and jostling for position on the bank.
There are patrol vehicles along the Mara now and the enforcement is tighter. We are glad about it. Those incidents damage the animals and they damage the profession.
For you as a traveller it comes down to three things.
One. Travel outside the peak if your dates allow it. Late July and October in the north often give you crossings as good as the second half of August with about half the vehicles. June at the Grumeti is quieter than anything that comes after it. And February at Ndutu is the strongest month of the year for predator action, at a lower price.
Two. Take more days in the north than you think you need. Booking one day at Kogatende is booking a chance, not a crossing. Three days is the least I would plan for.
Three. Choose a guide who can hold his position. At a crossing, a guide who stays where he is and switches the engine off is the difference between footage you keep and a traffic jam of people photographing the same thing. We know every guide who drives for us personally, and we chose them ourselves.
At the Mara River our guide stayed calm and held his position while everyone else drove off, expecting the herd to cross further along. The herd came past right in front of our vehicle, and the others had to turn back. Stephanie, September 2026
How we plan a migration trip
We start from your dates rather than from a standard route. You tell us when you can travel, and we tell you where the herds are likely to be, what you can realistically expect to see in that window, and where it is quiet enough to watch it properly. If your dates are poor for the migration, we say so, and we build the trip around something else that is strong in those weeks.
We live here and we have driven this route for seventeen years. The animals do not keep appointments, and we do not keep to schedules that pretend otherwise.
Download the migration itinerary, nine days along the route
It comes by email within a few minutes, and it has three things in it.
- The route day by day, from Tarangire and the Ngorongoro Crater up to the northern Serengeti.
- An indicative price per person, per season.
- Which trip length suits what you want to see of the migration.
Send us your date and we will set the itinerary against it, and against where the animals are that year.
