AbstractQuantitative and qualitative comparisons are made between laboratory measurements of rotating annulus flows and corresponding numerical model simulations. Two laboratory annuli, of similar dimensions but differing in instrumentation, are used. One contains a thermocouple array for temperature measurement: the other contains no sensor array but the working fluid is seeded with minute neutrally buoyant beads (600 m̈m diameter) which enable the horizontal velocity field to be measured. Each annulus has a rigid insulating lid in contact with the working fluid. the numerical model is a finite difference formulation based on the Navier‐Stokes equations for baroclinic flow of a Boussinesq liquid. Although the atmosphere and the laboratory annulus are both rotating baroclinic fluid systems, theforcing processesacting in the annulus are much simpler than those acting in the atmosphere, and may be accurately represented by established formulae: under a wide range of conditions no parametrizations of subgrid‐scale dynamical and diabatic processes are required. Comparison of numerical model results with laboratory measurements therefore enables the explicit dynamical formulation of numerical models of rotating, baroclinic flow to be verified to an extent which would be very difficult, if not impossible, to achieve using atmospheric data. Detailed quantitative comparisons for a steady wave flow reveal good agreement for major features of the temperature and horizontal flow fields, although a significant discrepancy in total heat flux is found. Qualitative comparisons are made by investigating the ability of the numerical model to reproduce the main flow types and phenomena of the laboratory system. Numerical simulations of intransitivity, hysteresis, wavenumber transitions, amplitude vacillation and a weak structural vacillation are described. Several suggestions for further comparative studies are made in conclu