Inter-annual climatic variability poses a substantial management and profitability challenge for pasture-based dairy producers in southern Australia. The effects of a range of seasonal scenarios on the production and profit of non-irrigated dairy farm systems using several different forage bases were investigated for two regions in southeast Australia using a systems modeling approach. For the Terang district, seasonal scenarios were constructed around combinations of early, average or late autumn rains, and short, average or long spring flushes. For the higher-rainfall Ellinbank district, scenarios were constructed around either above- or below-average summer–autumns or winter–springs.Compared to the ‘Base’ system where 100% of the grazing area was in perennial ryegrass (Lolium perenne), alternative systems such as double cropping and use of a summer shoulder (more summer-active) pasture based on tall fescue (Festuca arundinacea) returned similar or higher profit for most of the seasonal scenarios investigated. Only at Terang when the autumn rains were early and the spring was extended within the same year did the alternatives lose out to perennial ryegrass in predicted profit, reflecting the well-known excellent performance of perennial ryegrass in environments with a long growing season. Within each of the forage systems simulated, the profit outcomes were generally more consistent in the face of seasonal perturbations for the double cropping and summer shoulder pasture forage options than for those based on ryegrass. This suggests they are less prone to business risk. It reflects the apparent consistency across years of the yield of winter cereal during the period April–October when there is generally only small departure from ‘average’ conditions for plant growth between years. It also reflects the deeper effective rooting habit of tall fescue, and its ability to access more stored water in the soil profile compared to perennial ryegrass and thus greater capacity to ‘ride out’ periods of variability.Pasture growth analyses using climate data from 1900 to 1999 inclusive showed that above- or below-average conditions for plant growth in autumn and spring occur at frequencies between two and four years out of 10. Thus, all combinations of poorer than average or better than average seasons can be expected in any decade. This investigation suggests that diversification of the forage base will help smooth out between-year variability in profit provided the right forage types are selected, managed to a very high level, and effectively integrated within year-round feeding systems to achieve high milk production.