Abstract: Objective: Chlorophyll is an important photosynthesis pigment in crops. It not only reflects the photosynthetic capacity of the plant, but also indicates plant growth and nutritional status. Accurate acquisition of leaf chlorophyll content is very useful for monitoring crop growth and predicting crop yield. The portable optical instrument SPAD-502 can acquire the SPAD value quickly and in a non-destructive way, which indicate the leaf relative chlorophyll content. However, this approach requires a tremendous amount of hand labor and is not able to estimate leaf SPAD value in a large area, which did not meet the needs for monitoring the dynamic of crop growth and agricultural management. Optical remote sensing is an alternative for non-destructive measuring of leaf SPAD value at large scale. However, there are few studies to evaluate the effects of spectral resolution on optical remote sensing of winter wheat leaf SPAD value. Methods: Canopy spectral reflectance and leaf SPAD values of winter wheat were obtained under 4 growth stages (jointing, heading, anthesis and filling) and 3 nitrogen application levels through a 4-year field experiment. The leaf SPAD estimation model was constructed by using 25 commonly used chlorophyll content sensitive spectral indices combined with machine learning algorithms. The effects of five spectral resolutions on the reflectance of a single band, spectral indices and the estimation of SPAD value using machine learning algorithm were evaluated. Results and discussion: The sensitivity of single band reflectance to SPAD is affected by spectral resolution, growth stages and nitrogen application levels. In the whole growth period, the reflectance of red band is the more sensitive to SPAD than other bands, and the determination coefficient R² is between 0.411 and 0.579 at the 5 spectral resolutions. It was mainly due to the strong absorption of chlorophyll in the red band. At the most sensitive spectral resolution for red band in each growth stages, the R² was between 0.242 and 0.700, and the Var was between 0.313 and 0.952. The red edge band reflectance at 710 nm had the largest spectral resolution sensitivity variation coefficient, and Var was 1.000 in the whole growth period. The main reason is that the reflectance of green vegetation changes sharply in the red edge band, and there are obvious differences for the reflectance at different spectral resolutions. The spectral resolution affected the sensitivity of spectral indices to SPAD at different growth stages and nitrogen application levels. Except for the filling and anthesis stage, the spectral resolution range of the optimal sensitivity spectral index is between 25 and 50 nm. The potential reason is that for the optimal sensitivity spectral index, the broad spectral resolution makes the spectral index contain more spectral information and reduces the influence of noises. However, the effects of spectral resolution on the sensitivity of the optimal spectral index to SPAD is limited compared with that of single band reflectance. In the whole growth period, the spectral index mND705 presented the strongest sensitivity to SPAD at 50 nm (R² = 0.685) and a low variation coefficient of spectral resolution sensitivity (Var = 0.01). In each growth stage, the R² of the most sensitive spectral index to SPAD was between 0.387 and 0.895, and the coefficient of sensitivity variance Var was between 0.01 and 0.05. The increase of nitrogen application level enhanced the sensitivity of spectral index to SPAD. The main reason is that the improvement of nitrogen application level increases the leaf chlorophyll content and improve the ability of spectral index to detect SPAD. Optimizing spectral resolution of spectral indices as feature inputs for machine learning models improves the estimation accuracy of SPAD. In the whole growth period, the model constructed with 25 nm spectral resolution spectral index and PLSR had the best estimation accuracy, with R² of 0.82 and RMSE of 4.04. In each growth periods, the model constructed with the optimized spectral resolution spectral index as the input had the best estimation accuracy with R² between 0.53 and 0.92 and the RMSE was between 2.76 and 4.47. At the three nitrogen application levels, the R² was between 0.78 and 0.88 and the RMSE was between 2.67 and 4.55 using the model constructed with the optimized spectral resolution. The increase of nitrogen application level increased the chlorophyll content of leaves, and thus enhanced the detection ability of SPAD by the machine learning model based on spectral characteristics. Conclusion: This study investigated the spectral resolution effects on winter wheat leaf SPAD estimation. It provides a theoretical support for the design of optical remote sensing sensor, the development of spectral index and the optimization of spectral resolution to improve the estimation accuracy of winter wheat leaf SPAD using optical remote sensing.