Axisymmetric Drop Shape Analysis (ADSA) is a powerful technique for the measurement of interfacial properties from the shape of drops/bubbles. It relies on the best fit between theoretical curves with known surface tension values and an experimental profile. Despite the general success of ADSA, inconsistent results are obtained for nearly spherical drop shapes. Since the source of this and possible other limitations are unknown, the entire ADSA technique including hardware and software is systematically scrutinized. The hardware consists of electronics, and optical and mechanical components that generate a digital image of a drop. Since the quality of images has a considerable impact on the surface tension measurements, general guidelines for the use of hardware components are developed to enhance the quality of the image. The scrutiny of the software of ADSA is significantly more involved. The software consists of image analysis and numerical schemes. One of the key elements is the modularization of the software, since a generic software package is not suitable for all experimental situations. As a result, a more versatile image analysis module is introduced. In this context a variety of state-of-the-art edge detection techniques are studied, and a more robust technique is adopted. The two existing ADSA numerical schemes are also compared systematically, and the more efficient one is implemented. It is shown that even this superior numerical scheme has convergence problems for nearly spherical drops. This difficulty is due to numerical truncation and accumulation of round-off errors, which are the ultimate limitation of all numerical schemes. This intrinsic limitation becomes more pronounced as drops become closer to spherical in shape, but there were no objective criteria available to define "close to spherical drops". Therefore, a quantitative criterion called shape parameter is introduced to identify the range of applicability of ADSA. The improved version of ADSA not only determines the interfacial properties more accurately but, through the shape parameter, also provides an a priori knowledge of the accuracy of the results.