The experiment was conducted in the calcareous soil of Regional Agricultural Research Station (RARS) farm, Jessore, Bangladesh for 3 years from 2003- 2004 to 2005-2006. The land belongs to High Ganges River Floodplain agroecological zone and Gopalpur soil series (Soil taxonomy: Aquic Eutro- chrepts). The soil had high pH value (8.2) with low DTPA-Zn content (0.58 mg kg-1). The critical level of DTPA extractable Zn for maize in calcareous soils, as determined by Lindsay and Norvell (1978), is 0.78 mg kg-1. The other soil properties were 1.44% organic matter, 7.32 mg kg-1 Olsen-P, 0.29 c mol kg-1 exchangeable K, 20 c mol kg-1 Ca, 1.25 c mol kg-1 Mg, 7.51 mg kg-1 CaCl2-S and 0.21 mg kg-1 Ca (H2P04)2-B. Soil pH was determined by glass electrode pH meter (1:2.5 soil-water ratio) and organic matter by wet oxidation method (Nelson and Sommers 1982). The K, Ca and Mg contents of soil were determined by 1 M NH4OAC (pH 7.0) extraction method. The experiment was laid out in a randomized complete block design with three replications, each plot size being 5x4.5 m. Maize, mungbean and rice crops were grown in sequence. Zinc was applied to maize at 0, 2 and 4 kg ha-1 and its residual effect was evaluated on the following mungbean crop. The transplant aman rice (wet season rice) received Zn at 0, 1 and 2 kg ha-1. Thus, the number of treatments for the first, second and third crops was 3, 3 and 18 (3 x 2x3), respectively. Of the 18 treatments, 9 treatments (3x3) were with mungbean residue incorporation and 9 without mungbean incorporation. The total number of plots was 54 (3 x 2 x 3 x 3). This layout was kept undisturbed for the second and third years of the study. Zinc was supplied as ZnO (78% Zn). Every year the maize crop had received 260 kg N, 63 kg P, 80 kg K, 47 kg S and 1 kg B per hectare, and the rice had 80 kg N, 10 kg P and 10 kg S per hectare. No any nutrients were added for mungbean crop. The sources of nutrients were urea, TSP, MoP, gypsum and boric acid for N, P, K, S and B, respectively. Intercultural operations viz. weeding, irrigation and insecticide spray were done as and when required. Every year the maize (cv. Pacific 984, Thai hybrid) was sown in the first week of November and harvested in the first week of April. Thereafter mungbean (cv. BARI mung5) was sown in a week or so and harvested at the end of June. Rice (cv. BRRI dhan33) was transplanted in the second week of July and harvested in the second week of October. Data on the yield and yield contributing characters from all plots were recorded. The yield data were expressed as t ha-1 on 12% moisture basis. The grain and stover/straw samples from each plot for every crop were chemi- cally analyzed for N, P, S and Zn concentrations. Microkjeldahl method (H2S04 digestion) was fol- lowed for N determination and the HNO3-HCIO4 (5:1) digestion was made for P, S and Zn determinations. Nitrogen concentration was determined by titration method, the P and S concentration by colorimetric method and the Zn concentration directly by flame-AAS (Model UNICAM 969, England). The Zn uptake was calculated from the crop yield and Zn concentration data. Another field experiment was carried out at the same farm (RARS, Jessore) with split-plot design to test the response of some maize varieties (hybrids and composites) to added zinc, over 3 years (2002-2005). The hybrids were BARI hybrid maize 1, BARI hybrid maize 3, BARI hybrid maize top 1 and Pacific 984, and the composite varieties were BARI maize 6, Mohor, Barnali and Khoibhutta. There were two rates of Zn application, 0 and 3 kg ha-1 from ZnO. Thus, the total number of plots for this trial was 48 (8 varieties x 2 Zn treatments x 3 replications). Like crop- ping pattern experiment, every plot received an equal amount of N, P, K, S and B to sustain normal plant growth. Intercultural operations viz. weeding and irrigation were done whenever required. Every year the maize seeds were was sown in the first week of November and harvested in the first week of April. The grain yields from all plots were recorded and expressed as t ha-1 on 12% moisture basis. The data from both cropping pattern and single crop experiments were statistically analyzed follow- ing the principle of F-statistics and the mean values were separated by Duncan's Multiple Range Test (DMRT).