Abstract. An XMM-Newton observation of the cool (kT = 2.1 keV) cluster A1983, at z = 0.044, is presented. Gas density and temperature profiles are calculated over the radial range up to 500 h −1 50 kpc, corresponding to ∼0.35 r 200 . The outer regions of the surface brightness profile are well described with a β-model with β = 0.74, but the central regions require the introduction of a second component. The temperature profile is flat at the exterior with a slight dip towards the centre. The total mass profile, calculated from the temperature and density information assuming hydrostatic equilibrium, is consistent with an NFW profile, but with a low concentration parameter c = 3.75 ± 0.74, which may be due to the cluster not being totally relaxed. Published optical data are used to calculate the M/L B ratio profile and the overall iron mass over luminosity ratio. The M/L B ratio profile shows that, at large scale, light traces mass to a reasonable extent, and the M/L B ratio at 0.35r 200 is consistent with the trends with mass observed in the optical. The iron mass over luminosity ratio is about two times less than that observed for a cluster at 5 keV. The gas mass fraction rises rapidly in the central regions to level off quickly at ∼200 h −1 50 kpc; the value at 0.35 r 200 is ∼8%. The scaling properties of the emission measure profile are consistent with the empirical relation M gas ∝ T 1.94 ; use of the standard self-similar relation M gas ∝ T 1.5 results in a scaled profile that is a factor of about two too low as compared to the reference mean profile for hot clusters. Comparison of the entropy profile of this cool cluster with that of the hot cluster A1413 shows that the two profiles are extremely well scaled using the empirically determined relation S ∝ T 0.65 , suggesting that the slope of the S -T relation is shallower than expected in the standard self-similar model. The form of the two entropy profiles is remarkably similar, and there is no sign of a larger isentropic core in the cooler cluster. These data provide powerful agruments against preheating models. In turn, there is now increasing observational support for a trend of f gas with system mass, which may go some way towards explaining the observed scaling behaviour.