Characterizing and managing nutrient variability has been the focus of precision agriculture research for decades. Previous research has indicated that in-situ fluorescence sensor measurements can be used as a proxy for nitrogen (N) status in plants in greenhouse conditions employing static sensor measurements. Indeed, practitioners of precision N management require determination of in-season plant N status in real-time at field scale to enable the most efficient N fertilizer management system. The objective of this study was to ascertain if mobile in-field fluorescence sensor measurements can accurately quantify variability in maize canopy early in the crop growing season. This study was conducted across three site-years in 2012 and 2013 crop growing seasons. Multiplex®3, fluorescence sensor (Force-A, France) was used to collect plant N status measurements corresponding to the V6 and V9 maize growth stages. Conventionally, several fluorescence channels and especially derived indices are employed as predictors in a multiple linear regression analysis strategy to estimate plant nitrogen. These predictors are often cross-correlated among each other, which makes the regression analysis challenging. Hence, the new generation of experiments often leans towards machine learning strategies. Despite these regression approaches and their competitiveness, quantitative plant nitrogen estimation from fluorescence measurements using the independent channel information or various indices offers an opportunity to explore potential strategies with acceptable inversion accuracies. In this current study, fluorescence indices measured at V6 and V9 stages of maize were utilized for recommendations of selecting machine learning strategies among: (1) Multiple linear regression (MLR), (2) Support Vector Regression (SVR), (3) Gaussian Process Regression (GPR), (4) Random Forest Regression (RFR), and (5) Artificial Neural Network (ANN) Multi-layer perceptron. The preliminary results indicated that machine learning techniques outperform traditional workflow. The comparative analysis indicated a promising accuracy in estimation of plant N content, above-ground biomass, and N uptake at V6 stages of maize with the moderate range of correlation coefficient (r = 0.72±0.03) and Root Mean Square Error (RMSE). Indeed, the V9 stage results in better retrieval accuracies than V6 maize growth stage. Among the machine learning models, the Support Vector Regression (SVR) performed best for the 3 site-year data sets with a reasonable ranges of error estimates and yielding the lowest RMSE (0.36 and 0.23 (%N); 3.82 and 12.37g (biomass); 8.29 and 32.63g (N uptake) for V6 and V9 maize growth stages, respectively) for all three crop growth parameters.