Longitudinal data are collected in the repeated measurements format. Since those outcomes are correlated for the same subject, there are special statistical methods to handle the dependence between measurements. The objective of the thesis is to review the theory of Longitudinal data analysis and to investigate the statistical methods to analyze Longitudinal data. For the statistical methods, we focus on repeated measures models for normal and non-normal data. Moreover, we have applied the statistical meth ods on real life data of agriculture. The analysis of variance (ANOVA) for repeated measures (parametric approach), and the Friedman test (nonparametric approach) are used in the applications. The statistical analyses revealed that the factor time has a significant impact on the variables of interest in this study. Furthemore, the comparisons between different fertilizers are described in two different methods. The first one is the empirical method for comparing the performance of the fertilizers using a group of different variables. The second method is the one which is commonly used in Agriculture, for comparing different fertilizers. Both methods provide different results in the identification of the best fertilizer.