Accurate calculation of the time of melt onset, freeze onset, and melt duration over Arctic sea-ice area is crucial for climate and global change studies because it affects accuracy of surface energy balance estimates. This comparative study evaluates several methods used to estimate sea-ice melt and freeze onset dates: (1) the melt onset database derived from SSM/I passive microwave brightness temperatures (T_bs) using Drobot and Anderson's [J. Geophys. Res. 106 (2001) 24033] Advanced Horizontal Range Algorithm (AHRA) and distributed by the National Snow and Ice Data Center (NSIDC); (2) the International Arctic Buoy Program/Polar Exchange at the Sea (IABP/POLES) surface air temperatures (SATs); (3) an elaborated version of the AHRA that uses IABP/POLES to avoid anomalous results (Passive Microwave and Surface Temperature Analysis [PMSTA]); (4) another elaborated version of the AHRA that uses T_b variance to avoid anomalous results (Mean Differences and Standard Deviation Analysis [MDSDA]); (5) Smith's [J. Geophys. Res. 103 (1998) 27753] vertically polarized T_b algorithm for estimating melt onset in multiyear (MY) ice (SSM/I 19V–37V); and (6) analyses of concurrent backscattering cross section (σ°) and brightness temperature (T_b) from OKEAN-01 satellite series. Melt onset and freeze onset maps were created and compared to understand how the estimates vary between different satellite instruments and methods over different Arctic sea-ice regions. Comparisons were made to evaluate relative sensitivities among the methods to slight adjustments of the T_bcalibration coefficients and algorithm threshold values. Compared to the PMSTA method, the AHRA method tended to estimate significantly earlier melt dates, likely caused by the AHRA's susceptibility to prematurely identify melt onset conditions. In contrast, the IABP/POLES surface air temperature data tended to estimate later melt and earlier freeze in all but perennial ice. The MDSDA method was least sensitive to small adjustments of the SMMR–SSM/I inter-satellite calibration coefficients. Differences among methods varied by latitude. Freeze onset dates among methods were most disparate in southern latitudes, and tended to converge northward. Surface air temperatures (IABP/POLES) indicated freeze onset well before the MDSDA method, especially in southern peripheral seas, while PMSTA freeze estimates were generally intermediate. Surface air temperature data estimated latest melt onset dates in southern latitudes, but earliest melt onset in northern latitudes. The PMSTA estimated earliest melt onset dates in southern regions, and converged with the MDSDA northward. Because sea-ice melt and freeze are dynamical transitional processes, differences among these methods are associated with differing sensitivities to changing stages of environmental and physical development. These studies contribute to the growing body of documentation about the levels of disparity obtained when Arctic seasonal transition parameters are estimated using various types of microwave data and algorithms.