Annular cores were formed in start-up core physics tests of the High Temperature Engineering Test Reactor (HTTR) to obtain experimental data for verification of design codes. The first criticality, control rod (CR) positions at critical conditions, neutron flux distribution, excess reactivity, etc., were measured as representative data. These data were evaluated with the MVP Monte Carlo code, which can consider directly the heterogeneity of coated fuel particles (CFPs) distributed randomly in fuel compacts. It was made clear that the heterogeneity effect of CFPs on keff's for annular cores is smaller than that for fully loaded cores. The measured and the calculated keff's agreed with each other with differences <1%k. The calculated neutron flux distributions agreed with the measured results. A revised method was applied for evaluation of excess reactivity to exclude the negative shadowing effect of CRs. The revised and calculated excess reactivity agreed with differences <1%k/k.