Pot experiment to investigate the performance of upland rice in a previously alang-alang dominated area was 
conducted under greenhouse condition at BIOTROP, Bogor, Indonesia from November 1986 to May 1989. The 
treatments were  factorially  combined,  replicated  5  times  and  randomized  completely.  The  first  factor  was 
alang-alang  control  consisting  of  5 different  techniques,  i.e.  (1)  glyphosate  applied  at  2.2  kg  a.e./ha;  (2) 
imazapyr applied at 1.5 kg a.e./ha; (3) dalapon applied twice at 7.4 + 7.4 kg a.i./ha; (4) slashing followed by 
soil cultivation; (5) slashing of alang-alang only; while the second factor was nitrogen fertilizer at 4 different 
levels, i.e. (1) 0 kg N/ha, (2) 60 kg N/ha, (3) 120 kg N/ha given twice, 60 kg N/ha at planting time and 60 kg 
N/ha at 38 dap, (4) 180 kg N/ha given twice, 90 kg N/ha at planting and 90 kg N/ha at 38 dap. 
Plant height (cm), tiller number/pot, productive tiller (%), panicle length (cm), spikelets/panicle, empty 
spikelet (%), weight 1000 grains (g) and grain yield (ton/ha) were observed. 
Upland rice grown with zero tillage technique using glyphosate (2.2 kg a.i./ha) or dalapon (14.8 kg a.i./ha) 
performed as good as or even better than manual cultivation. Imazapyr at 1.5 kg a.e./ha was phytotoxic to 
rice planted 1 month after spraying. The application of N fertilizer lower than 60 kg N/ha was not sufficient, but 
more than 60 kg N/ha was too high; it stimulated the production of too many tillers, with high percentage of 
unproductive tillers and empty grains.