Hundreds of millions of people suffer an annual cycle of hunger and hardship that is linked to the agricultural season. There is a growing argument that it should theoretically be possible to design social protection measures that are counter-cyclical such as employment schemes, or social transfer programmes that only operate at specific times of the year. Conceptually, this is all very neat. But the reality is different for a number of reasons. The first problem is that seasonality is not quite as clear-cut as its proponents imply. The second is that, on top of the general fluctuation of the seasons, individual households are additionally exposed to a range of non-seasonal income and expenditure, which may cumulatively have as great or greater impact on their individual resources and resilience. Third, there are other cycles – both longer and shorter than seasonality – that impact on households. To design social protection responses that could adequately address all such cycles, and the complex interplay between them, would make such programmes impossibly complicated to administer. And there is a danger that getting it wrong might inflict additional damage, by undermining traditional social support mechanisms, and fuelling the social and economic polarisation that can accompany seasonality. Two possible, but as yet unproven, mechanisms are discussed that attempt to address seasonality through more flexible social protection: employment guarantee schemes and seasonal risk reduction. But the key message is that it is much better to provide a simple, regular, predictable and consistent transfer throughout the year, and to allow individual households to do their own budgeting and husbanding of resources, than to make paternalistic assumptions of need based on an overly simplistic conceptualisation of the vagaries and impacts of seasonality.