A nonlinear least squares program (MULTI) for microcomputers was developed for pharmacokinetic analysis. Written in BASIC, it uses four algorithms—Gauss-Newton, damping Gauss-Newton, modified Marquardt, and simplex—for nonlinear curve fitting. Up to five pharmacokinetic equations can be simultaneously fitted to observed time courses. The program was demonstrated using ampicillin and oxacillin data in humans. The program allows users to define pharmacokinetic equations, which are then fitted to data. The simplex method is robust and widely used in chemical engineering. The residual sum of squares (SS) and Akaike's information criterion (AIC) are used to evaluate model fit and complexity. AIC is defined as N ln(SS) + 2M, where N is the number of data points and M is the number of parameters. Input data is handled through conversational BASIC features, and output includes estimated parameters, SS, AIC, and predicted values. The simplex method does not calculate standard deviations due to its different principle. The program can be adapted for different microcomputers with minor changes. The two-compartment model equation is an example of a pharmacokinetic equation defined in the program. The program was tested with artificial data and real human data, showing good convergence. The inclusion of lag time improved model fit for ampicillin excretion. The program was also used for oxacillin and its metabolites, demonstrating its versatility. The simplex method showed good convergence even with poor initial parameter estimates, while Gauss-Newton methods were more efficient with close initial estimates. The minimum AIC estimation may suggest model selection but cannot detect flip-flop models. The program is a valuable tool for pharmacokinetic analysis on microcomputers.A nonlinear least squares program (MULTI) for microcomputers was developed for pharmacokinetic analysis. Written in BASIC, it uses four algorithms—Gauss-Newton, damping Gauss-Newton, modified Marquardt, and simplex—for nonlinear curve fitting. Up to five pharmacokinetic equations can be simultaneously fitted to observed time courses. The program was demonstrated using ampicillin and oxacillin data in humans. The program allows users to define pharmacokinetic equations, which are then fitted to data. The simplex method is robust and widely used in chemical engineering. The residual sum of squares (SS) and Akaike's information criterion (AIC) are used to evaluate model fit and complexity. AIC is defined as N ln(SS) + 2M, where N is the number of data points and M is the number of parameters. Input data is handled through conversational BASIC features, and output includes estimated parameters, SS, AIC, and predicted values. The simplex method does not calculate standard deviations due to its different principle. The program can be adapted for different microcomputers with minor changes. The two-compartment model equation is an example of a pharmacokinetic equation defined in the program. The program was tested with artificial data and real human data, showing good convergence. The inclusion of lag time improved model fit for ampicillin excretion. The program was also used for oxacillin and its metabolites, demonstrating its versatility. The simplex method showed good convergence even with poor initial parameter estimates, while Gauss-Newton methods were more efficient with close initial estimates. The minimum AIC estimation may suggest model selection but cannot detect flip-flop models. The program is a valuable tool for pharmacokinetic analysis on microcomputers.