Knowledge on the growth of rice (Oryza sativa L.) in low-fertility field conditions is essential to ensure their sustainability and enhance productivity. The key variables determining the productivity of such systems were studied in 40 recommended rice varieties grown in a low-fertile field. The paddy field had received no form of fertiliser or straw for the past 40 years, but it was used for rice cultivation two times per year under other standard crop-management practices. Harvests were made at 6 weeks after planting and at physiological maturity. The same varieties were also grown for 14 days in glass boxes containing distilled water only. Stepwise regression was used to identify the important variables for predicting dry weight (DW) at 14 days, 6 weeks and maturity. Thousand-seed weight and seed phosphorus (P) content of seed paddy (grains used for planting) had strong positive correlations with DW of seedlings after 14 days (r = 0.91, P < 0.0001) and 6 weeks (r = 0.7, P < 0.0001), respectively. However, initial growth performances did not correlate well with aboveground DW or grain yield at maturity. DW at maturity positively correlated with time taken to maturity (r = 0.78, P < 0.0001), shoot K content (r = 0.28, P = 0.008), rootDW(r = 0.22, P < 0.02), and uptake rates ofK(r = 0.32, P < 0.01) and P (r = 0.33, P < 0.01) at 6 weeks. Seed N concentration decreased (r =0.63, P < 0.001) and growth rate increased (r = 0.65, P < 0.001) with time taken to maturity. Although seed weight and seed P content of rice are key determinants of early crop establishment, capacity to produce large roots, amount of K taken up, P and K uptake rates at 6 weeks, and time taken to maturity are the key determinants of maturity biomass and grain yield in low-fertile field conditions. [ABSTRACT FROM AUTHOR]