The effects of three human premixes (Mixtard, Actraphane, Humulin M3), syringe mixed 30% regular and 70% NPH insulin, regular insulin alone and NPH insulin alone, on intermediary metabolism, plasma free insulin levels and action profiles were compared using the euglycemic clamp technique. Seven normal volunteers received 20 IU of each insulin subcutaneously in a randomized fashion on separate days. The first and last 60 min of the 6 h clamp were chosen as summary measures of clinical importance. Significantly elevated plasma free insulin levels were found with all treatments compared to NPH insulin alone during the first hour, although by the final hour only Mixtard produced significantly higher levels compared to NPH (19.4 +/- 1.2, 10.5 +/- 0.3 mU/l P less than 0.01, respectively). Analysis of area under the incremental insulin absorption curve demonstrated that Mixtard produced significantly increased levels compared to syringe-mixed regular: NPH (7.6 +/- 0.8), Actraphane (9.6 +/- 1.0) and Humulin M3 (9.0 +/- 0.8 mU/l all P less than 0.05). Mixtard also resulted in significantly higher glucose infusion rates compared to the other premixes. No difference in action was found between regular and pre- or syringe-mixed human insulins during the first hour of the studies. The effects on intermediary carbohydrate and lipid metabolism were similar for syringe and premixed insulins. We conclude that: (1) fixed human insulin mixtures with NPH cause no blunting of the action of the soluble component. (2) Actraphane and Humulin M3 are similar but Mixtard may have a greater effect on some aspects of insulin action. (3) In clinical practice, fixed human insulin mixtures will be as efficacious as syringe-mixed preparations but may be easier and more convenient to use.