Laboratory channels are often used to study the complexity of braiding mechanisms for the advantages with respect to field studies. Nevertheless, the extensive use of experimental data raises the question of how representative laboratory braided channels are as compared to real braided rivers. This study verifies to what extent laboratory braided patterns reproduce the main features of braided rivers. Experimental data display isotropic and anisotropic scaling of braided patterns, state-space plot of total widths, anisotropic scaling of islands, and statistical distribution of island areas that are similar to those observed in real rivers. Moreover, the data reveals scaling in the perimeter-area relation. These results support both the reliability of experimental braided channels as physical models of braided rivers and also the possibility of investigating some aspects of braiding in the laboratory that are difficult to address in the field. The lack of preferential scales in island characteristics suggests that other phenomena must play a key role in generating island shapes besides classical sediment transport-based mechanisms which tend to select well-defined length scales.