Detailed investigations of the temperature distribution in the Rhine Graben indicate a regular pattern of graben-parallel thermal anomalies following major north-south striking faults. These faults, which might imply north-south convection pattern due to higher permeability, have been ignored in conventional models based on 2D east-west striking sections. The present study investigates small-scale temperature anomalies in the Rhine Graben, focusing on its north-south striking major faults and their important permeability structures. Analytical solutions for convective flow in vertical faults were applied to a clearly observable anomalous temperature pattern. Through these calculations the fault geometry, minimal fault permeability and time to convective onset were derived from observed distances between anomalies. Since analytical solutions are limited to simple model geometries, further improvement was achieved through numerical model simulations, which allow the assumption of more complex initial and boundary conditions. Using the finite volume code TOUGH2 we investigated in detail the convective fluid flow pattern along the Γ-fault in the Landau region through a 3D model. There, regions of alternatively higher and lower temperatures follow each other at an average distance of 1800 m. Focusing on the predominant north-south permeability structure, our model consists of a vertical north-south striking fault and surrounding matrix with fault geometries based on analytical predictions. The observed temperature could be explained by a numerical model indicating fluid velocities in the fault of ∼10 -9 m s -1 and an anomaly whose minimal age is 77,000 years. The study proves that the observed small-scale graben-parallel temperature anomalies can be explained only by convection systems within north-south striking fault zones. The regional flow regime is therefore significantly influenced by the convective fluid flux in major north-south striking faults. Investigations of temperature anomalies in the Rhine Graben must take into account this north-south flow. Since the presented results correlate with findings from other geothermal anomalies in the Rhine Graben such as the one at Soultz, we suspect this to be a general feature of the major north-south striking faults there. Moreover, since in most grabenlike structures thermal anomalies are associated with the Graben master faults, the present study may provide common clues for investigating other graben-parallel temperature anomalies.