Cast-in-situ short paneled concrete pavements (CiSPCP) are a class of innovative concrete pavements, which are being considered as a sustainable replacement for jointed plain concrete pavement. However, there is a lack of understanding of various aspects of CiSPCP such as joint spacing, long-term joint performance, dominant stresses, failure criteria, and curling behavior. In this study, the curling behavior of CiSPCP test sections with different slab sizes and thicknesses on National Highway (NH)-18 (old NH-33) were investigated during the summer of 2019 and the winter of 2020. The vertical displacements at mid-slab (longitudinally and diagonally to the slab), the effect of slab size and thickness on curling with seasonal variation were considered, along with a comparison of measured curling using theoretical analysis. The measured displacements at the edge of slabs were smaller compared with those at the center of the slab. Interestingly, the occurrence of maximum slab curl was not in tandem with the maximum temperature gradient (TG). The time lag for the response was around 2 to 4.5 h of the occurrence of a maximum TG. This observation was very significant because temperature stresses have a profound effect on stresses in concrete pavement. However, this field observation has indicated that the lag in the development of maximum TG may also lead to a lag in the development of stresses. Further, it was found that the theoretical method overestimates the curling compared with the field-measured displacements, which can be mainly attributed to the lag.