The quantitative effect of different preparation variables on the sodium content of cooked dry pasta was evaluated. Semolina spaghetti (< 5 mg sodium/100 g) was cooked by a typical method (454 g, 5.68 L water, 36 g salt, al dente, no rinsing) and after systematic variation of amount of salt, water:pasta ratio, cooking volume and time, rinsing, pasta shape, whole grain. Sodium was assayed by ICP-MS, including rigorous quality control. Pasta cooked without salt had < 5 mg sodium/140 g serving, and 247-490 mg/serving when cooked in salted water by the different variations. Rinsing reduced sodium by 34%. There was a linear relationship between salt concentration in cooking water and sodium in cooked pasta; doubling the concentration increased sodium by 243 mg/serving (> 100% of 2300 mg/day), relative to the reference method. No other variables affected sodium. Results allow more accurate estimation of sodium intake from cooked pasta, since food composition tables that do not reflect variations in cooking parameters.