Zinc (Zn) oxide is the most widely used fertiliser for the predominantly acidic to neutral soils of south-western Australia. For these soils, the residual value of Zn oxide has been determined for wheat and lupin, but not for barley, oats, canola and triticale, which are also grown in the region. Just after termination of a long-term (17 year) field experiment that measured the residual value of Zn oxide for wheat, soil samples were collected from selected plots to use in 2 glasshouse experiments. The field experiment was on previously unfertilised, newly cleared duplex soil (sand with much lateritic ironstone gravel over clay) and before the experiment started DTPA extractable Zn for the top 10 cm of soil was <0.2 mg Zn/kg. In the first glasshouse experiment, soil samples from the nil-Zn treatment of the field experiment were used to measure the critical Zn concentration in young mature growth of 6 crop species (wheat, barley, oats, lupin, canola and triticale) when 7 levels of Zn (0, 50, 100, 150, 200, 300 and 600 mg Zn/kg soil) were applied to the soil. In the field experiment, 0.5 and 1.0 kg Zn/ha, as Zn oxide, had been applied once only in each of the following years to previously nil-Zn plots: 1983, 1984, 1986, 1990, 1992, 1996 and 2000. Soil samples were collected from these plots to use in the second glasshouse experiment. This experiment estimated how long the Zn treatments applied in the field remained effective, as estimated using shoot yields and critical Zn concentrations in young mature growth of the same 6 crop species used in experiment 1. Critical Zn concentration in young mature growth was about 14 mg/kg for wheat, barley and lupin, 15 mg/kg for triticale, and 18 mg/kg for oats and canola. The residual value of Zn varied with crop species. As estimated from shoot yields, the 0.5 kg Zn/ha treatment was effective for ≤10 years for wheat, barley and oats, ≤14 years for lupin and canola, and >17 years for triticale. The 1.0 kg Zn/ha treatment remained fully effective for all crop species. As determined from projected estimates of the data, the time taken for Zn concentrations in young mature growth to reach critical values, the residual value of the 0.5 and 1.0 kg Zn/ha treatments were least for wheat, barley and oats, were greater for lupin and canola, and greatest for triticale. There were a total of 7 wheat crops and 10 pasture years during the 17 years of the field experiment. For the 0.5 and 1.0 kg Zn/ha treatment applied in the field in 1983, 30–34% of the applied Zn was removed in grain of the 7 wheat crops grown before soil samples were collected to do the glasshouse experiments. The pasture was grazed by sheep and it was estimated that 16–24% of the Zn applied in 1983 may have been removed in wool and meat. Removal of Zn in grain and animal products therefore decreased the residual value of the Zn oxide fertiliser.