It is well documented that multi-species assemblages are needed to maintain a high-functioning turfgrass ecosystem. However, evaluation and analysis of turfgrass seed mixtures has historically relied upon methods that provide little or no predictive capability to identify the optimal mixture of constituent species. The objectives of this work were to: 1) introduce the use of simplex designs for mixture experiments in the design of seed mixtures for turfgrass communities, and 2) demonstrate the identification of optimal seed mixtures through the application of basic mixture experiment methodology to a simple 4-component seed mixture analysis. Fifteen mixtures of Kentucky bluegrass (Poa pratensis L.), hard fescue [Festuca trachyphylla (Hack.) Krajina], perennial ryegrass (Lolium perenne L.), and tall fescue (Festuca arundinacea Schreb.) were chosen based on simplex designs for mixture experiments and four replications were established from seed in 20.95 cm diameter pots within a climate-controlled growth chamber. Mixtures were fertilized weekly and clipped to 5 cm every 7-19 days, with the final clipping occurring 19 wk after seeding. Total dry clipping weight was used as a response variable to which response surface and numerical optimization methods were applied to identify an optimal seed mixture. Results of the analysis showed total dry clipping weight would be maximized for a mixture consisting of 83% Lolium perenne and 17% Festuca trachyphylla, which was not part of the initial trial design. The experiment was repeated with the addition of this optimal mixture. Comparison of the optimized mixture to the design mixtures revealed it to be second best in the second experimental run, and best for the combined data set of both runs. Results show that the application of simplex designs and numerical optimization to seed mixture experiments has great potential to optimize turfgrass community health; however, variability in mixture performance indicates the need for further assessment of the method.