Esophageal balloons of three different wall thicknesses, two different lengths, and two different diameters were made to assess which type of balloon gave the most consistent and reliable measurements of dynamic compliance (CL) and pulmonary resistance (Rp). The balloons were subjected to in vitro testing to determine their pressure-volume characteristics and working range and then used in infants to compare in vivo results from one balloon to another. The optimal balloon was found to have a length of 35--50 mm, a diameter of 7.6 mm, and a wall thickness of 0.045--0.075 mm. The use of unsuitable balloons or inappropriate volumes of air within the balloon resulted in applied pressures being under-recorded, both in vitro and in vivo, with consequent overestimation of CL and underestimation of Rp during infant lung function tests. Recommendations are made concerning the techniques of making and using esophageal balloons.