Urease from dehusked seeds of watermelon was immobilized in 1.5% agarose gel with 53.9% entrapment. There was negligible leaching(<10% at $4^{circ}C$) and the same gel membrane could repeatedly be used for seven days. The immobilization exhibited no apparent change in the optimum pH but there was a significant decrease in the optimum temperature ($50^{circ}C$ as compared to $65^{circ}C$ for soluble urease). The immobilized urease revealed an apparent $K_m;of;9.3{pm}0.3 M;$ 1.2 times lower than the soluble enzyme $(11.4{pm}0.2mM)$. Unlike soluble enzyme which was inhibited at 200mM urea, the immobilized urease was inhibited at 600mM of urea and above, and about 47% activity was retained at 2M urea. The time-dependent thermal inactivation kinetics at 48 and $52^{circ}C$ was found to be biphasic, in which half of the initial activity was destroyed more rapidly than the remaining half. These gel membranes were also used for estimating the urea content of the blood samples from the University hospital. The results obtained matched well with those obtained by the usual method employed in the clinical pathology laboratory. The significance of these observations is discussed.