The island is small enough that most planning tools treat it as one climate. It is not. A ridge of hills down the middle splits the rainfall so thoroughly that the department maintains separate zone classifications for either side of it, and the practical consequence is that the same week can be the best and the worst time to travel depending on which shore you booked.

The four seasons, as the department defines them

The Department of Meteorology's own Climate News Letter sets out the year in four parts: the first inter-monsoon from March to April, the southwest monsoon from May to September, the second inter-monsoon from October to November, and the northeast monsoon from December to February. These are the official labels, and the department's forecasts and warnings are issued against them rather than against a wet-and-dry binary.

The two monsoons are not symmetrical in what they touch. The southwest monsoon brings heavy rain to the southwestern lowlands and the western slopes of the hill country. The northeast monsoon brings rain mainly to the northern and eastern parts of the island. The two inter-monsoon periods behave differently again, delivering rain more broadly rather than to one flank. Our guide to how long crossing the island takes covers that separately.

Yala and Maha are not the monsoons

This is worth stating plainly because the error is widespread. Yala and Maha are cropping seasons, and each one spans two meteorological seasons rather than corresponding to a single monsoon.

The department's own journal defines them: Maha covers the second inter-monsoon from October to November together with the northeast monsoon from December to February, and Yala covers the first inter-monsoon from March to April together with the southwest monsoon from May to September. Anything that equates Yala with the southwest monsoon alone, or uses the two words as monsoon names, has compressed a four-season calendar into two and lost the inter-monsoon months in the process.

What the district figures actually say

The thirty-year means for 1991 to 2020 make the coastal split legible without any interpretation. For September, October and November, Colombo records 277.6, 382.0 and 350.3 millimetres. Jaffna over the same three months records 62.3, 210.2 and 334.4. Trincomalee records 97.7, 205.8 and 335.2.

September is the number that carries the decision. Colombo's 277.6 millimetres against Jaffna's 62.3 is more than a fourfold difference in the same month on the same island, and it is the tail of the southwest monsoon doing it. By November the three converge into the 330s, which is the second inter-monsoon reaching everywhere at once, and the reason October and November are poor months to bet on any single coast.

The zone classification follows the same logic. The wet zone is the southwest, the portion directly exposed to the southwest monsoon; the rest of the island falls into the intermediate and dry zones. The department's mapped annual rainfall across the island ranges from 1,062.4 to 4,612.1 millimetres, a spread of more than four to one within a country you can cross in a day.

Which months the hazards cluster in

Rainfall totals and disaster risk are not the same question, and the historical record separates them usefully. On the DesInventar disaster profile, floods are most likely in May, at the opening of the southwest monsoon, and again in December with the northeast monsoon. Landslides concentrate later, in November, December and January. The broader disaster peaks fall across April to June and October to December.

The warning system is three-colour and worth recognising on sight. Red means take action, amber means be prepared, green means the threat is over. Rainfall is banded separately as fairly heavy at 50 to 100 millimetres, heavy at 100 to 150, and very heavy above 150. Sea condition warnings come from the department's Natural Hazards Early Warning Centre rather than from a fisheries authority, with an advisory issued for winds of 55 to 65 kilometres per hour and a warning for winds up to 70. There is a companion piece on the Poya calendar.

The current season is not a normal one

The department's seasonal forecast issued on 4 September 2026 projects above-normal rainfall for September to November in the Western and Sabaragamuwa provinces and in the Galle, Matara, Kandy and Nuwara Eliya districts, with below-normal September across most of the country and above-normal November nationally. It attributes this to a strengthening El Niño, with a greater than ninety per cent chance of a very strong event.

That forecast contradicts an earlier one. The department's southwest monsoon release of 12 May 2026 gave slightly below-normal rainfall for the same districts over an overlapping period. The later document is the one to work from, but the reversal is the point: a seasonal forecast for an El Niño year is a weaker planning instrument than the thirty-year means, and should be read as a flag rather than a schedule.

What the visitor numbers already encode

Arrivals track the northeast monsoon's effect on the southwest coast almost exactly. The peak months are December, January and February, precisely when the southwest is sheltered, and the trough is May and June, when the southwest monsoon opens. The 2025 monthly report puts the low at 132,919 in May against 258,928 in December.

The inference is narrower than it looks. Those numbers describe when visitors go to the island's southwestern resorts, not when the island as a whole is at its best. Travelling the east coast in July, or the southwest in January, means travelling with the crowd; the reverse means travelling against both the weather and the pricing. The adjacent question of how the island differs from India is treated separately.

What this guide does not cover

Several figures were deliberately omitted because no official source could be verified for them. The millimetre thresholds that formally separate the wet, intermediate and dry zones are not stated here, nor are the National Building Research Organisation's numeric landslide triggers, nor a national annual rainfall average. Official whale-watching months and surf season months are likewise absent; the marketing claims in circulation for both could not be traced to a government or scientific publisher.

This guide also gives no sea temperatures, no per-beach swimming safety assessments, and no forecast beyond the department's own published seasonal window. For anything time-critical, the Department of Meteorology's current warnings are the operative document, not a thirty-year mean.