<p>In this paper, the soil temperature dynamics in south-western Nigeria is investigated using nonlinear time series analysis approach. The 5-mins interval soil temperature data observed continuously over a period of 1 year (from January to December, 2012) were analysed to investigate the existence of chaos. We employed the method of false nearest neighbour carefully interlaced with the average mutual information method to reconstruct the original attractor of the soil temperature's evolution process. The Lyapunov exponent, in addition to the strange attractor of the process, was also employed to identify chaos in the process. Numerical simulations were performed to show the validity of our approach. Tabular and graphical are presented. The results of the computation show the existence of chaos in the soil temperature.</p>