We have developed a nondestructive assay of the fule content of deuterium‐tritium (DT) ‐filled microballoon laser fusion targets, which is based on &bgr;‐particle counting rates. Using a model employing transmission measurements of kilovolt electrons through thin films, observed count rates are correlated with the amount of tritium in the glass walls and hollow interior of the microballoons. It has been shown that gas pressure in balloons can be calculated from tritium content and a knowledge of the initial gas composition since D and T were found to leak at the same rate. This assay technique is primarily applicable for balloons with glass wall thicknesses <1.5 &mgr;m where the number of escaping &bgr; particles is large compared with the number of x‐ray photons generated in the glass. The technique has been applied to measure the pressure retention characteristics of individual targets. At room temperature, the balloons exhibited widely diverse and rapid leakage rates which could not be correlated with a model based on molecular diffusion and the assumption that all balloons had a homogeneous composition. Cryogenic storage greatly reduced the leakage rates with pressure retention half‐lives ranging from 5 to approximately 12 years. Finally, it was determined that tritium in the glass wall is trapped and evidence is presented to support the hypothesis that it is uniformly distributed across the shell.