Mass and concentration of clusters of galaxies are related and evolving with redshift .
We study the properties of a sample of 31 massive galaxy clusters at high redshift , 0.8 \mathrel { \lower 2.58 pt \hbox { $ \buildrel \textstyle < \over { \scriptstyle \sim } $ } } % z \mathrel { \lower 2.58 pt \hbox { $ \buildrel \textstyle < \over { \scriptstyle \sim } $ } } 1.5 , using weak and strong lensing observations .
Concentration is a steep function of mass , c _ { 200 } \propto M _ { 200 } ^ { -0.83 \pm 0.39 } , with higher-redshift clusters being less concentrated .
Mass and concentration from the stacked analysis , M _ { 200 } = ( 4.1 \pm 0.4 ) \times 10 ^ { 14 } M _ { \odot } / h and c _ { 200 } = 2.3 \pm 0.2 , are in line with theoretical results extrapolated from the local universe .
Clusters with signs of dynamical activity preferentially feature high concentrations .
We discuss the possibility that the whole sample is a mix of two different kinds of haloes .
Over-concentrated clusters might be accreting haloes out of equilibrium in a transient phase of compression , whereas less concentrated ones might be more relaxed .