AbstractThe paper presents the results of an original experimental program that was focused on punching shear behavior of flat slabs supported by rectangular columns with a column rectangularity index cmax/cmin = 1.0, 4.33, and 6.33 and slabs supported by elongated columns that were loaded by four different load patterns. Seven isolated flat slab specimens with a thickness of 200 mm were tested until their failure. An effect of the column geometry and the way of the flat slab loading on punching shear capacity of slab‐column connections were investigated. The obtained experimental results were confronted with predictions carried out with the design provisions from code of practice EC2 (2004), ACI 318‐19, and prEC2 (2021) and the original Critical Shear Crack Theory (CSCT) (2008) model developed by Muttoni. Experimental results confirmed an effect of column rectangularity on punching shear resistance when a value of shear strength ratio Vtest/Vmodel was below one in the specimens with cmax/cmin > 2. The highest punching resistance was achieved in slab specimen with loading pattern x/2y where x is an axis along the column dimension cmax. The lowest capacity was achieved with a loading pattern 9x/y where 90% of the load flew perpendicular to the axis “y” parallel with cmin. The most accurate and balanced predictions provided CSCT (2008) model.