We present results from the flight of two prototype CZT detectors on a scientific balloon payload in September 2000. The first detector, referred to as ``CZT1,'' consisted of a 10 mm x 10 mm x 2 mm CZT crystal with a single gold planar electrode readout. This detector was shielded by a combination of a passive collimator surrounded by plastic scintillator and a thick BGO crystal in the rear. The second detector, ``CZT2,'' comprised two 10 mm x 10 mm x 5 mm CZT crystals, one made of eV Products high pressure Bridgman material and the other of IMARAD horizontal Bridgman material, each fashioned with a 4 x 4 array of gold pixels on a 2.5 mm pitch. The pixellated detectors were flip-chip-mounted side by side and read out by a 32-channel ASIC. This detector was also shielded by a passive/plastic collimator in the front, but used only additional passive/plastic shielding in the rear. Both experiments were flown from Ft. Sumner, NM on September 19, 2000 on a 24 hour balloon flight. CZT1 recorded a non-vetoed background level at 100 keV of ~1e-3 cts/cm2/s/keV. Raising the BGO threshold from 50 keV to ~1 MeV produced only an 18% increase in this level. CZT2 recorded a background at 100 keV of ~4e-3 cts/cm2/s/keV in the eV Products detector and ~6e-3 cts/cm2/s/keV in the IMARAD detector. Both CZT1 and CZT2 spectra were in basic agreement with Monte Carlo simulations, though both recorded systematically higher count rates at high energy than predicted. No lines were observed, indicating that neutron capture reactions, at least those producing decay lines at a few 100 keV, are not significant components of the CZT background. Comparison of the CZT1 and CZT2 spectra indicates that passive/plastic shielding may provide adequately low background levels for many applications.