National Ignition Facility (NIF) specifications require nondestructive, independent profiling of copper, argon, and oxygen in a delivered beryllium capsule. We use a combination of two methods to accomplish this goal: (a) model-enhanced energy dispersive spectroscopy (EDS) of witness shell fragments for destructive profiling of all three elements in a select sample within a batch and (b) differential radiography (DR) to profile copper and argon on multiple shells to nondestructively prove the sample-to-sample consistency within a batch. This combination fully qualifies the delivered shells. For EDS, we developed a physics model and fabricated standards to quantify low concentrations of relatively light elements in a very low-Z matrix. For model validation, we produced sputtered beryllium capsules containing a single dopant in each shell and used contact radiography (CR) to characterize the dopant profiles to 5 to 10% accuracy. The copper calibration was also checked against bulk Cu-Be standards with known composition, and the argon and oxygen calibrations were also checked against the X-ray absorption edge spectroscopy (Edge method) and the weight gain methods. Together, the EDS method gives ±0.1, ±0.05, and ±0.2 at.% accuracy for copper, argon, and oxygen, respectively, in NIF specification capsules. For DR, we conduct two CR measurements with the X-ray tube running at 9 and 30 kVp, respectively. The differential response between copper and argon enables elemental separation. The dopant profiles can be measured to ±0.1 at.% for NIF specification capsules. The oxygen profile in DR must be inferred from the EDS measurement. In the production work flow, we use EDS to obtain the oxygen profile and use it as input to the DR measurement. We then check that the copper and argon profiles obtained from DR and those from EDS are consistent. The average argon and copper contents from either method can also be checked against the results from the Edge method. These two levels of cross-checks offer critical assurances to the data integrity in production metrology.