The topic of peritectic phase transition in steel has been widely studied using various in-situ experimental techniques including visual observation on a surface of a sample in a high temperature microscope and measurement of the thermal response from the bulk of the sample using differential thermal analysis (DTA) and differential scanning calorimetry (DSC). These two different techniques have inherent limitations leading to difficulties in interpretation of the results in the case of complex phase transformations such as the peritectic reaction, and there have been discrepancies among the interpretations from many previous studies of the peritectic transition in steel. A new technique was developed through a spatial combination of a DTA apparatus and a high-temperature laser-scanning confocal microscope to overcome these limitations. Solidification experiments with carbon steels of 0.061, 0.182 and 0.451 wt%C were conducted under conditions typical of conventional thermal analysis experiment, i.e. 90 mg of sample mass and 10 ��C/min cooling rate, and the findings of earlier studies were revisited. We concluded that whether or not a peritectic reaction in a steel specimen of peritectic composition occurs, largely depends on the size of the sample used in the experiment. Deep undercooling for austenite nucleation can be experienced, in the absence of the occurrence of a peritectic reaction as a result of constrained nucleation of austenite, due to a limited number of nucleation sites in the small samples used.