We plan to test the substellar nature of a selection of brown dwarf
 candidates in young (age under 10 Myr), embedded clusters. We detected
 them in JHK surveys of the cores of Rho Ophiuchi and NGC 2024. They
 are the best brown dwarf candidates in these clouds, identified on the
 basis of a new method of analysis that fits the near IR photometry
 simultaneously to theoretical models and interstellar reddening. The
 method separates the effects determining the observed spectral energy
 distribution (SED) of stellar objects and provides a mass estimate,
 which ultimately determines the stellar or substellar nature. For ISO
 observations, we selected objects indicated by our fits to lie well
 below the end of the main sequence (with mass 0.03-0.07 solar masses).
 We also rejected objects with strong IR excesses indicative of large
 amounts of circumstellar material, studied by some Guaranteed Time
 programme proposals. We want to determine the SED from 3 to 8 microns.
 Together with the near IR points, these new data can tightly constrain
 the luminosity of these objects, with temperatures of 2500 K-3000 K.
 The ISO data will either detect the photosphere or identify IR excess
 due to circumstellar material. As masses are most tightly constrained
 by luminosities, not near infrared colors, ISO data will provide a
 critical test of whether these objects are truly young brown dwarfs.
 Together with our ground based observations, we will have sufficient
 data to accurately determine temperature and luminosity, and hence
 masses, of the objects. Broad band photometry with CAM will accurately
 constrain the shape of the SED. We plan to observe in four bands in
 which there is valuable information unavailable from the ground. SW1
 samples mostly the stellar photosphere. LW1 and LW2 provide an
 extended wavelength baseline for sampling the SED at high sensitivity.
 This permits a reliable disentangling of the factors causing IR excess
 in the observed SED. LW4 gives a more accurate measurement of the
 circumstellar contribution for the brightest objects.