Human NPH U100 (100 IU/ml), given once daily, is often absorbed too fast to cover the basal insulin demand throughout 24 h. The aim of the present study was to examine whether the absorption of human insulin suspensions would be delayed by increasing the insulin concentration from U100 to U500. In each experiment 10 IU of corresponding human Ul00 and U500 preparations, labelled with 125I-insulin, were injected subcutaneously and contralaterally in the neck of a pig followed by monitoring residual radioactivity over the injection sites. The time until 75, 50 and 25% residual radioactivity (T(75%), T(50%) and T(25%)) using NPH U500 was compared with NPH U100 in 14 experiments: T(75%): 5.0+/-0.5 (mean+/-SEM) vs 3.8+/-0.3 h (P=0.007, paired t-test), T(50%): 12. 2+/-0.9 vs 9.0+/-0.6 h (P=0.003) and T(25%). 24.2+/-1.2 vs 17.9+/-1. 2 h (P=0.001). The corresponding values for semilente U500 compared with semilente U100 in eight experiments were: T(75%): 2.8+/-0.4 vs 1.6+/-0.2 h (P=0.02), T(50%): 5.6+/-0.6 vs 3.4+/-0.3 h (P=0.01) and T(25%): 10.9+/-1.1 vs 7.2+/-0.7 h (P=0.009). Thus, the absorption of a given dose of human NPH or human semilente insulin in pigs is substantially delayed by changing the insulin concentration from Ul00 to U500. Human NPH U500 appears to be more appropriate than human NPH U100 for injection once daily in basal insulin therapy.