In this paper we develop five new methods of estimation to estimate the parameters of four widely used nonlinear models namely Haldane, Powell, Moser and Webb model. A standard growth data set of Escherichia Coli is considered for estimating the parameters. The estimat-ed model parameters are analyzed by evaluating statistical parameters χ2, Root Mean Square
Error, R2, R2a and R2pre. As a result, the Powell model gives the best fit with estimation of R2 as 99.7% with respect to method IV. Moreover, the other three models also provide remarkable fit along with the newly introduced methods. Method II gives R2 value as 99% in case of the Haldane model. The method IV estimates with R2 value as 99.6% in the Moser model and the method III estimates with R2 value as 99.4% in case of the Webb model.