Rainfed farms in south-eastern Australia often combine annual cropping and perennial pasture phases with grazing sheep enterprises. Such diversity serves in managing diseases, pests and plant nutrition while stabilising income in the face of wide, uncorrelated variations in international commodity prices and local weather over time. We use an actuarial accounting approach to capture the above contexts to render financial risk profiles in the form of distributions of decadal cash balances for a representative 1,000 ha farm at Coolamon (34°50′S 147°12′E) in NSW, Australia. It is for the soil and weather conditions at this location that we pose the question of which is best: establishing the perennial pasture, lucerne (Medicago sativa L.), either sown with the final crop of the cropping phase, or alone following the final crop? It is less expensive to sow lucerne with the final crop, which can provide useful income from the sale of grain, but can reduce pasture quantity and quality in poorer years. Though many years of field research have confirmed sowing lucerne alone is the most reliable way to establish a pasture in this area, and years of extension messages to this effect have gone out to farmers, they often persist in sowing lucerne with their final cereal crops. For this region, counting all costs, we show that sowing lucerne alone can reduce farm financial risk (probability of negative decadal cash balances) at stocking rates above 10 dse/ha, compared to the practice of sowing lucerne with a cover crop. At low stocking rates (i.e., 5 dse/ha), there appears to be no financial advantage in either establishment approach. We consider the level of equity, background farm debt and overhead costs to demonstrate how these also affect risk profile positions of the two sowing options. For a farm that is deeply in debt, we cannot suggest either approach to establishing lucerne will lead to substantially better financial outcomes.