The emergent universe scenario is obtained in a flat universe with the equation of state (EoS) ( $$p=B\rho -A\rho ^{\frac{1}{2}}$$ ) (where A and B are constants) as per Mukherjee et al. Here observed Hubble data (OHD) and binned Pantheon data of supernovae are used to determine the best-fit EoS parameters of the emergent universe models. Evolution of deceleration parameter q and EoS $$(\omega =\frac{p}{\rho })$$ has been studied through their plots with redshift. The statefinder diagnostics consisting of jerk parameter j, snap parameter s and Om diagnostics are introduced to make a comparative study of the emergent universe models. The parameters j, s have been plotted with redshift, with themselves, with q, to examine the evolution of the universe and its possible future in these models. Present values of q, $$\omega $$ , j, s along with their cosmological averages have been determined for the emergent universe models $$(B=-\frac{1}{3},0,\frac{1}{3},1)$$ . Akaike information criterion (AIC) and Bayesian information criterion (BIC) are the two model selection tools used here, according to it $$B=\frac{1}{3}$$ is the better model, then the $$B=0$$ model. Squared adiabatic sound speed $$c_{s}^{2}$$ is required to be positive in order that structure formation is a reality. Models with $$B=1,\frac{1}{3}$$ show clear signs and $$B=0$$ model shows hints of structure formation in the past. However, $$B=1$$ model is not compatible with the observational data. The $$B=-\frac{1}{3}$$ model does not show any signs of structure formation $$(c_{s}^{2}<0)$$ , so is plagued with instability. Best-fit values of the two suitable emergent universe model with (i) $$B=0$$ are $$H_{0}=69.29$$ , $$A_{s}=0.61$$ , $$q(0)=-0.42$$ , $$\omega (0)=-0.61$$ , $$j(0)=0.47$$ , $$s(0)=0.20$$ and (ii) $$B=\frac{1}{3}$$ are $$H_{0}=70.21$$ , $$A_{s}=0.77$$ , $$q(0)=-0.54$$ , $$\omega (0)=-0.69$$ , $$j(0)=0.75$$ , $$s(0)=0.08$$